UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

On ambiguous sentences with the particle To Fujimura, Taiji 1977

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1977_A8 F84.pdf [ 4.05MB ]
Metadata
JSON: 831-1.0094169.json
JSON-LD: 831-1.0094169-ld.json
RDF/XML (Pretty): 831-1.0094169-rdf.xml
RDF/JSON: 831-1.0094169-rdf.json
Turtle: 831-1.0094169-turtle.txt
N-Triples: 831-1.0094169-rdf-ntriples.txt
Original Record: 831-1.0094169-source.json
Full Text
831-1.0094169-fulltext.txt
Citation
831-1.0094169.ris

Full Text

& • I ON AMBIGUOUS SENTENCES WITH THE PARTICLE TO by TAIJI FUJIMURA B.A., International Christian University, 1973 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OP THE REQUIREMENTS FOR THE DEGREE OP MASTER OF ARTS i n THE FACULTY OF GRADUATE STUDIES (Dept. of Asian Studies) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA March, 1977 © T a i j i Fujimura, 1977 In presenting th is thes is in p a r t i a l fu l f i lment of the requirements for an advanced degree at the Univers i ty of B r i t i s h Columbia, I agree that the L ibrary shal l make it f ree l y ava i lab le for reference and study. I fur ther agree that permission for extensive copying of th is thes is for scho la r l y purposes may be granted by the Head of my Department or by h is representat ives . It is understood that copying or pub l i ca t ion of th is thes is for f i n a n c i a l gain sha l l not be allowed without my wr i t ten permission. Department of Asian Studies The Univers i ty of B r i t i s h Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Date A p r i l 14, 1977 i i ABSTRACT This thesis attempts to describe Japanese sentences in which to functions as the symmetric, coneomitative, or coordinate particle. Earlier studies by Gkutsu (1967), Kuno (1967-68) and Inoue (1976b) are closely examined. It is assumed that sentences with conjoined HP's like Taroo to Hanako ga kekkon-si-ta are a l l transformationally derived. In connection with this assumption, the rules of Conjunction Reduction, Issyo ni Insertion, NP Scrambling in NP, Conjunct Movement and Reciprocal Transformation are proposed. It is also argued that the rules should apply in the order given above. i i i TABLE OF CONTENTS ACKNOWLEDGEMENTS V INTRODUCTION . . . . . . . 1 0.1 1 0.2 1 Chapter 1: PRELIMINARY ANALYSIS OF DATA AND RECENT STUDIES . . . . . . 3 1.1 3 1.1.1 4 1.1.2. "i 6 1.1.3 9 1.1.3.1 10 1.1.3.2 12 1,1,3.3 15 1.1.4 17 1.2. 19 1.2.1 19 1.2.2 22 1.2.3 • 30 Chapter 2: DERIVED CONJUNCTION 35 2.1 35 2.2 35 2.2.1. 35 2.2.2 39 2.2.3. ' • 42 2.3 44 iv 2.3.1 44 2.3.2 48 2.3.3 50 2.3.4 56 2.4 . • . 67 2.4.1 67 2.4.2 70 2.4.2.1 71 2.4.2.2 74 2.4.2.3 75 Chapter: CONCLUDING REMARKS 77 3.1 77 3.2 77 FOOTNOTES 81 BIBLIOGRAPHY 88 V ACKNOWLEDGEMENTS I would l i k e to express my deepest gratitude to Professor Matsuo Soga f o r his constant encouragement and invaluable c r i t i c i s m s and suggestions. I would also l i k e to thank Brenda Sawada f o r her assistance i n editing. In addition, I wish to thank the Department of Asian Studies f o r providing me with an opportunity to act as a teaching assistant during the academic years 1975-77. And f i n a l l y , I am grateful to my wife Mami f o r her patience and assistance as an informant. INTRODUCTION 0 . 1 . The descriptive study of a language has the construction of a grammar as i t s goal. There have been constructed a great number of Japanese grammars. However, most of them are deficient i n that they do not r e f l e c t the native speaker's knowledge of his language ( i . e . , competence). Although these grammars generate a l l the observed gram-matical sentences of the language, they do not provide correct pre-dictions of sentences not observed. In other words, such grammars are not descriptively adequate, even i f they are observationally adequate. Their inadequacy seems to derive from the fact that they leave untouched (or unexpressed) many of the underlying r e g u l a r i t i e s of sentences. 0 . 2 . Recent development of the theory of transformational grammar seems to show that i t i s the optimal theory f o r describing competence (e.g., i n t u i t i o n about the relationship between sentences and so on). Within this theory, to describe the syntactic aspects of sentences, a grammar contains two types of rules: context-free phrase structure rules, which generate underlying (or deep) structures; and grammati-cal transformations, which map underlying structures onto surface structures. The intuitions about the relationship between sentences are captured by deriving related sentences from the same or highly similar underlying structures by means of transformations. 2 This t h e s i s attempts to c l a r i f y the s t r u c t u r e of Japanese sentences with the p a r t i c l e to b a s i c a l l y w i t h i n the framework of the theory proposed by Chomsky (1965). The main concern here i s to determine the appropriate underlying structures f o r to sentences, and to formulate grammatical transformations f o r the d e r i v a t i o n of well-formed surface s t r u c t u r e s . / 3 Chapter 1 PRELIMINARY ANALYSIS OF DATA AND RECENT STUDIES 1.1. The particle to_ can be used in various kinds of sentences as the following examples shpw:^ (1) Taroo ga Hanako to kekkon-si-ta. with marry do PAST 'Taro married Hanako.• (2) Taroo ga Hanako to gakkoo e i t - t a . school to go •Taro went to school with Hanako.' (3) Taroo to.Hanako ga kekkon-si-ta. 'Taro and Hanako married,* (4) Taroo to.Hanako ga gakkoo e i t - t a . 'Taro and Hanako went to school.' (5) Basu to „torakku to takusii ga syoototu-si-ta. "bus.. truck. taxi collide 'The bus, the ..truck, and. the taxi collided. • (6) Zyon ga i s y a t o nat-ta. John doctor to become. 'John became a medical doctor.* (7) Zyon ga rainen Nihon.e ik-u.to i t - t a . next, year Japan go say * John said that he would go to Japan next year.' (8) Zyon wa byooki-na noni, gakkoo e ik-oo to.,.si-ta. i l l though try •John tried to go to school, ..though he was i l l . ' (9) Zyon ga uti ni kaer-u to sugu, tomodati kara denwa ga kakat-home to return immediately friend from phone te ki-ta. come 'As soon as John returned home, he had a phone call.from a friend.' However, my concern here is with sentences like (l),-(5.) only. These to!s are called the symmetric particle in (1), the concomitative particle in (2), and the coordinate particle in (3.)-(5,). .1.1.1... In logic a relation R,is .called, symmetric, i f for each, R-.pa.ir the relation R also holds in the inverse direction. In other words, a relation ..R, is called symmetric .such that i f xRy is true,. then yRx is also true. According to.this definition, we could say that the verbs kekkon-si in (1) and i t in. (.2) both .belong to the group, of predicates which express a symmetric relation, for we.have, the f o l -lowing sentences, which ,are always true i f sentences (1) and.(2) are true. (10) Hanako. ga. Taroo to kekkon-si-ta. 'Hanako married Taro.* , (11) Hanako ga Taroo to gakkoo e i.t-ta. •Hanako went to school with Taro.* However, those predicates are quite different syntactically. The difference shows up in sentences which.do not take a noun phrase followed_ by the particle, to., (12) Taroo ga kekkon-siTta. !Taro.married (someone).' (13) Taroo.ga, gakkoo e it-:ta. 'Taro went to school.' As the English,translations show, sentence (12) implies that,there was someone to whom Taro got married, but sentence (13) does not imply that there was someone with whom Taro went to school. In other words, kekkon-si must take a noun phrase followed by the symmetric p a r t i c l e to (henceforth, a symmetric NP) i n the deep s t r u c t u r e e v e n i f that NP i s deleted at some stage of derivation, while i t i n (13) does not have such a co-occurrence r e s t r i c t i o n . Thus the underlying structure of sentence (12) would be (14) Taroo ga dareka to kekkon-si-ta, someone while that of sentence (13) would be very simi l a r to i t s surface structure. Such predicates l i k e kekkon-si are often ca l l e d "symmetric predicates". Since a symmetric predicate can co-occur with a noun phrase followed by the concomitative p a r t i c l e to (henceforth, a concomitative NP), sentence (1) can have two readings: (1) Taroo ga Hanako to kekkon-si-ta. a. 'Taro got married to Hanako.' b. 'Taro got married (to someone) together with Hanako.' If Hanako i s Taro's s i s t e r , (b) i s the most natural reading, showing that they married different spouses at a double wedding. In this reading sentence (1) can be paraphrased as (15) Taroo ga Hanako to issyo n i kekkon-si-ta. together 'Taro got married (to someone) together with Hanako.' Note that sentence (15) i s not ambiguous. Prom now on, when a sen-6 tence (e.g., sentence (x)) has more than two readings (e.g., a, b, and c ) , I s h a l l r e f e r to sentence (x) i n the sense of (a) reading as sentence (x/a). 1.1.2. Compare sentence (2) with the sentence given below: (16) Taroo ga Hanako to i s s y o n i gakkoo e i t - t a . •Taro went to school together with Hanako.* As the English t r a n s l a t i o n shows, the paraphrase r e l a t i o n e x i s t s between sentences (2) and (16). However, these sentences behave d i f f e r e n t l y when they are embedded as a r e l a t i v e clause i n a noun phrase. In Japanese, the r e l a t i v e clause i s placed immediately before the NP i t modifies. In other words, an NP containing a clause has the form s NP^Jin the deep s t r u c t u r e . The surface s t r u c t u r e of the NP i s formed by simply d e l e t i n g the i d e n t i c a l NP i n the embedded sentence* Thus when Taroo of sentence (2) i s r e l a t i v i z e d ( i . e . , when sentence (2) i s embedded as a modifier to the NP Taroo), (17) r e s u l t s from the a p p l i c a t i o n of R e l a t i v i z a t i o n . (17) Hanako to gakkoo e i t - t a Taroo 'Taro who went to school with Hanako' The underlying structure of (17) would be that shown i n (18). (18) ^ N P ^ S T a r 0 ° g a H a n a k o t o S a k k 0° e i t - t a J Taroo} In the d e r i v a t i o n of (17) from ( 1 8 ) r R e l a t i v i z a t i o n a p p l i e s to (18) to delete the f i r s t occurrence of Ta.roo ( i . e . , Taroo i n the embedded sentence) under the i d e n t i t y with the 'head' NP Taroo. Then the dangling p a r t i c l e (ga i n (18)), which results from the deletion of Taroo. i s deleted by convention. When Taroo of sentence (16) i s r e l a t i v i z e d , (19) i s derived from the underlying structure (20) i n the same manner. (19) Hanako to issyo n i gakkoo e i t - t a Taroo 'Taro who went to school together with Hanako' (20) [^p {_ gTaroo ga Hanako to issyo n i gakkoo e i t - t a ^ Taroo]| Note that the paraphrase r e l a t i o n s t i l l holds between (17) and (19). Next, l e t us see what happens when the concomitative NP's ( i . e . , Hanako's) of sentences (2) and (16) are r e l a t i v i z e d . (21) Taroo ga gakkoo e i t - t a Hanako | XHanako who Taro went to school' (22) Taroo ga issyo n i gakkoo e i t - t a Hanako 'Hanako whom Taro went to school together with' The symbol x indicates that the sentences with i t are unacceptable. Only (22) i s acceptable, although the only difference between (21) and (22) i s that (22) contains issyo n i while (21.) does not. Their underlying structures would be those shown i n (24) and (25) respec-t i v e l y . (23) [ Np [gTaroo ga Hanako to gakkoo e i t - t a J HanakoJ (24) £ Np [gTaroo ga Hanako to issyo n i gakkoo e i t - t a ^ j Hanakojj The difference i n the grammaticality of (21) and (22) shows that a concomitative NP should not be deleted by Rela t i y i z a t i o n except when i t i s followed by the phrase issyo n i . 8 Sentence (2) can be paraphrsed not only by (16) but also by ( 2 5 ) . (25) Taroo ga Hanako to tomp n i gakkoo e i t - t a . with 1Taro went to school with Hanako.' Most native speakers would say that i s s y o n i i n (16) and tomo n i i n (25) are interchangeable, and that the former-is l e s s formal than the l a t t e r . But observe the f o l l o w i n g p a i r s of sentences: X. f a . issyo n i l Lb. tomb n i J ( (26) Taroo wa Hanako to 7 ^ c n o gakusei da. of student i s •Taro and Hanako are both students of UBC.' Ta. x i s s y o n i l (27) Sekiyu wa se k i t a n to "S V moe-yasui. petroleum coal '•b. tomo n i J burn easy 'Petroleum and coal are both flammable.' x. f a . issyo n i l (28) Eigo wa suugaku to < { z y u u y ° 0 de-ar-u. English mathematics ' b. tomo n i -> important i s 'English and mathematics are both important.' While (a) sentences of (26)-(28) are not acceptable, (b) sentences are a l l acceptable. Notice that the predicates of those sentences are a noun (+copula ), an a d j e c t i v e , and an a d j e c t i v a l noun (+copula) r e s p e c t i v e l y . These predicates w i l l be r e f e r r e d to as "state p r e d i -cates". The a c c e p t a b i l i t y d i f f e r e n c e between (a) sentences and (b) sentences seems to i n d i c a t e that tomo n i has two d i s t i n c t usages; one as a s t y l i s t i c v a r i a n t of issyo„ni as i n (25) and the other as a d i f f e r e n t l e x i c a l item from issyo n i , which should not co-occur with state predicates. Therefore, i t i s safe to say that sentence (25) has the f o l l o w i n g two readings: 9 (25) Taroo ga Hanako to tomo n i gakkoo e i t - t a . a. 'Taro went to school together with Hanako.' b. 'Taro and Hanako both went to school.' Note that sentence (2) cannot have the second reading of sentence (25). That i s , the paraphrase r e l a t i o n holds between sentence (2) and sen-tence (25/a), not between (2) and (25/b). The d i f f e r e n c e between issyo n i and tomo n i w i l l be examined again i n sees. 1.1.3.1-3 below. 1.1.3. Sentence (3) can have three readings: (3) Taroo to Hanako ga kekkon-si-ta.^ a. 'Taro and Hanako got married (to each oth e r ) . ' b. 'Taro and Hanako got married (to someone together).' c. 'Taro and Hanako got married (to someone sep a r a t e l y ) . ' Notice that sentence (3) takes a symmetric predicate and that i t contains a conjoined NP as i t s subject. Sentence (3/b) i s a para-phrase of sentence (1/b). Sentence (3/b) can be also paraphrased as (29), which i s not ambiguous. (29) Taroo to Hanako ga issyo n i kekkon-si-ta. •Taro and Hanako got married (to someone) together.' On the other hand, sentence (3/c) means only that Taro and Hanako married d i f f e r e n t spouses, and i t i s noncommittal as to whether they married together or not. This sentence can be paraphrased as (30). (30) Taroo to Hanako ga tomo n i kekkon-si-ta. 'Taro and Hanako both got married.' 10 Sentence (30) i s Unambiguous as i s sentence (29). On the other hand, sentence (4) can have two readings only, although i t seems to have the same kind of structure as sentence (3). (4) Taroo.to Hanako ga gakkoo e i t - t a . a. 'Taro and Hanako went to school (together).' b. 'Taro and Hanako (both) went to school.' The> kind of semantic diffe r e n c e between sentences (3) and (4) seems to stem from the f a c t that the former takes a symmetric predicate, while the l a t t e r does not. Sentence (4/a) and sentence (4/b) can be paraphrased as (31) and (32) r e s p e c t i v e l y . (31) Taroo to Hanako ga issyo n i gakkoo e i t - t a . 'Taro and Hanako went to school together.' (32) Taroo to Hanako ga tomo n i gakkoo e i t - t a . •Taro and Hanako both went to school.• Note that sentence (31) i s als o a paraphrase of sentence (16). 1.1.3.1. The semantic diffe r e n c e between iss y o n i and tomo n i can be seen more c l e a r l y i n the f o l l o w i n g dialogues: (33) A: }Taroo to Ziroo ga kekkon-si-ta-soo-da ne. I hear ' I hear Taro and J i r o got married. Is that r i g h t ? 1 B: Aa. Taroo to Ziroo wa tomo n i kekkon-si-ta yo. •Yes. Taro and J i r o both got married, I t e l l you.' 4 : Issyo n i s i - t a no ka. do Q • ( L i t . ) Did they do together?' B: Iya. Betubetu n i . s i . - t a - r a s i i yo. no separately seem 11 'No. ( L i t . ) They seem to have done separately, I t e l l you.' j (34) A: Taroo to Ziroo ga kekkon-si-ta-soo-da ne. (=(33A)) 'I hear Taro and J i r o got married. Is that r i g h t ? ' B: Aa. Taroo to Ziroo wa issyo n i kekkon-si-ta yo. •Yes. Taro and J i r o got married (to someone) together, I t e l l you.• M A : Tomo n i s i - t a no ka. ' ( L i t . ) Did they both do?' Since the f i r s t (B) sentence of (33) i s noncommittal as to the "togetherness" of Taro's and J i r o ' s marriages, the second (A) sen-tence i s a possible question. On the other hand, the (B) sentence of (34) i s not ambiguous as to that p o i n t , so the second (A) sentence i s a meaningless question. (The symbol # i n d i c a t e s that the sentence i s meaningless;) Consider the f o l l o w i n g p a i r s of sentences: ^a. issyo n i | (35) Eiga o mi-te, Tomoko to Yosiko wa < 7 s i s s i n - s i - t a . movie see 'b. tomo n i ' f a i n t , a. Tomoko and Yoshiko f a i n t e d together.i 'Upon seeing the movie, 4 r 'b. Tomoko and Yoshiko both f a i n t e d . ^ a. x i s s y o n i ; (36) Tomoko to Yosiko wa J [ betubetu n i syuusyoku-si-ta. '•b. tomo n i ' enter the serv i c e r a . xentered the serv i c e together separately.! •Tomoko and Yoshiko j f b. both entered the serv i c e separately. The phrase iss y o n i i n (35a) can be replaced by doozi n i 'at.the same time', but tomo n i i n (35b) cannot. Sentence (36a) i s semantically anomalous, while sentence (36b) i s i n t e r p r e t a b l e . The semantic anomaly of sentence (36a) i s due to the f a c t that the meaning of issyo n i i s 12 e n t i r e l y opposite to that of betubetu n i 'separately'. Therefore, a sentence without betubetu n i becomes p e r f e c t l y acceptable as shown i n (37). (37) Tomoko to Yosiko wa issyo n i syuusyoku-si-ta. 'Tomoko and Yoshiko entered the service together.' By the same token, sentences (38) and (39) are also acceptable, although they are d i f f e r e n t from sentence (37) i n meaning. (38) Tomoko to Yosiko wa tomo n i syuusyoku-si-ta. •Tomoko and Yoshiko both entered the s e r v i c e . ' (39) Tomoko to Yosiko wa betubetu n i syuusyoku-si-ta. •Tomoko and Yoshiko entered the service separately.' With respect to sentence (36b), some native speakers might not accept i t as r e a d i l y as (36a). I t seems to me that the u n a c c e p t a b i l i t y of ( 3 6 b ) — i f i t i s u n a c c e p t a b l e — i s due to the semantic redundancy between the two phrases tomo n i and betubetu n i . To conclude, sentences with tomo n i are noncommittal as to the togetherness of the actions which two p a r t i c i p a n t s perform (as i n (38)); however, that togetherness i s c l e a r l y expressed i n sentences w i t n issyo n i (as i n (37)). 1.1.3.2. I have mentioned i n sec.1,1.2 that issyo n i should not co-occur with some kinds of predicates ( i . e . , state p r e d i c a t e s ) . The co-occurrence r e s t r i c t i o n between issyo n i and those predicates s t i l l holds i n sentences which take conjoined NP's. 13 r a . issyo nil (40) 3atoo-san to Suzuki-san wa J" 7 Ninon-zin da. <-b. toino ni J Japanese •Mr. Sato and Mr. Suzuki are both Japanese..' x. ra. issyo ni-j (41) Satoo-san to Suzuki-san wa •< v se ga takai. l b . tomo ni J t a l l 'Mr. Sato and Mr. Suzuki are both t a l l . ' r a. xissyo n i l (42) Watakusi wa kono e to ano s ga < I su^^-I this picture that <-b. tomo ni J like I I like both this picture and that picture.' Therefore, we could conclude that a grammar of Japanese contains the following constraint: (43) State predicates should not co-occur with the phrase issyo ni in a simple sentence. This constraint implies that issyo ni may co-occur with verbs. This implication is confirmed by the fact that (a) and (b) sentences of (44)-(46) are a l l acceptable. (44) a. Tomoko to Yosiko wa ima made issyo ni koko ni i-ta. now until here at exist 'Tomoko and Yoshiko were here together until now.' b. Tomoko wa ima made Yosiko to issyo ni koko ni i-t a . •Tomoko was here together with Yoshiko until now.' c (45) a. Tomoko to Yosiko wa ima Amerika ni it - t e - i - r u . to go exist 'Tomoko and Yoshiko are in America together now.' b. Tomoko wa ima Yosiko to issyo ni Amerika ni it-t e - i - r u . 'Tomoko is in America together with Yoshiko now.' (46) a. Taroo wa biiru to uisukii o issyo ni non-da. beer whiskey drink 'Taro drank beer and whiskey together.1 14 b. Taroo wa biiru o uisukii to issyo ni non-da. 'Taro drank beer together with whiskey.' In regards to the, sentences in (45), one might claim that issyo ..ni co-occurs with i_t,'to go' in a simple sentence, which is embedded as a complement of an aspectual verb ter-i-^ru.., However, although the claim is accepted, those sentences are not contrary to our expectation, for it, is also a verb. Next, consider the following pairs of sentences: (47) a. Tomoko to Yosiko wa issyo ni UBC.no gakusei ni nat-ta. student become \n - , ) 'Tomoko and Yoshiko became students of UBC together.' I b. Tomoko wa Yosiko to issyo ni UBC no gakusei ni nat-ta. 'Tomoko became a student of UBC together with Yoshiko.' (48) a. 'Watakusi wa kono e to ano e ga issyo ni suki ni nat-ta. 'I became fond of this picture and that picture together.' ) f b. 'Watakusi wa kono e ga ano e to issyo ni suki ni nat-ta. •I became fond of this picture together with that picture.' In order that the constraint stated in (43) s t i l l apply to the sen-tences in (47), we must assume nat 'to become', takes a complement and that issyo ni co-occurs with nat-ta, not gakusei (+copula). in a simple sentence; otherwise, as the constraint (43) prevents issyo ni from co-occurring with a noun (+copula), those sentences would be regarded as ungrammatical. However, even i f we admit that issyo ni co-occurs with the verb nat, we cannot account for the awkwardness of the sentences in (48). To expain the awkwardness, a further examination of the constraint on sentences with issyo ,ni ..is needed, but i t is beyond the scope of the present study, ijor our purpose 15 we w i l l be t e n t a t i v e l y s a t i s f i e d by saying that issyo n i must co-occur with verbs. 1.1.3.3. I have stated i n sees. 1.1.2-3 that sentences without issyo n i can be synonymous with those with i t . In sec. 1.1.3, I have a l s o stated that sentences without tomo n i can be synonymous with those with i t . This l a t t e r statement s t i l l proves v a l i d , because the sen-tences below are ^ synonymous with (b) sentences of (40)-(42). (49) Satoo-san to Suzuki-san wa Nihon-zin da. 'Mr. Sato and Mr. Suzuki are both Japanese.' (50) ' Satoo-san to Suzuki-san wa se ga t a k a i . 'Mr. Sato and Mr. Suzuki are both t a l l . ' (51) Watakusi wa kono e ,to ano e ga s u k i da. •I l i k e both t h i s , p i c t u r e and that p i c t u r e . ' I t should be noted that these sentences take not only state predicates but als o conjoined NP's. However, compare the f o l l o w i n g sentences with (b) sentences of (26)-(28), which take state predicates only. (52) xTaroo wa Hanako to UBC no gakusei da. (53) Sekiyu wa se k i t a n to moe-yasui. (54) Eigo wa suugaku to zyuuyoo de-ar-u. These sentences are a l l unacceptable. Their u n a c c e p t a b i l i t y stems c l e a r l y from the f a c t that they do not have tomo n i . Therefore, as f a r as sentences containing tomo n i are concerned, we could say something l i k e t h i s : (55) Sentences without tomo n i can be synonymous with those with i t , as long as they take conjoined NP's.1 16 According to the statement given i n (55), the f o l l o w i n g p a i r s of sentences would have to be synonymous with each other. (56) a. Taroo wa gakkoo to kyookai e tomo n i i t - t a . church 'Taro went to both school and church.' b. Taroo wa gakkoo to kyookai e i t - t a . 'Taro went to both school and church.' (57) a. Taroo wa Tomoko to Yosiko to tomo n i kekkon-si-ta. 'Taro got married (to someone) together with Tomoko and Yoshiko.' b. Taroo wa Tomoko to Yosiko to kekkon-si-ta. 'Taro got married to Tomoko and Yoshiko.' The sentences i n (57) are not synonymous, though those i n (56) are. As the English t r a n s l a t i o n s show, sentence (57a) states that Taro got married to someone other than Tomoko and Yoshiko, but sentence (57b) states that he got married to them. I t should be noted, however, that (b) sentence of (58) can be synonymous with (a) sentence. (58) a. Taroo wa Tomoko to / Yosiko to tomo n i kekkon-si-ta. 'Taro got married to Tomoko together with Yoshiko.' b. Taroo wa Tomoko to / Yosiko to kekkon-si-ta. •Taro got married to Tomoko with Yoshiko.' Notice that these sentences do not take conjoined NP's. Therefore, to account f o r the meaning di f f e r e n c e between (a) and (b) sentences of (57), (55) should be changed i n t o (59). (59) Sentences without tomo n i can be synonymous with those with i t , as long as they take conjoined NP's but not symmetric predicates. 17 1.1.4. Let us consider sentences l i k e (5) i n which three NP's are conjoined and c o n s t i t u t e a subject. Sentence (5) i s three-way ambiguous l i k e sentence (3), although (a) i s the most n a t u r a l reading. (5) Basu to torakku to t a k u s i i ga s y o o t o t u - s i - t a . a. 'The bus, the truck, and the t a x i c o l l i d e d (with one another).' b. 'The bus, the truck, and the t a x i c o l l i d e d (with something together).' c. 'The bus, the truck, and the t a x i c o l l i d e d (with something separ a t e l y ) . ' The verb syoototu-si 'to c o l l i d e ' i s a symmetric predicate, because i f (a) sentence of (60) i s t r u e , (b) sentence i s als o true. (60) a. Basu ga torakku to s y o o t o t u - s i - t a . 'The bus c o l l i d e d with the truck.' b. Torakku ga basu to s y o o t o t u - s i - t a . 'The truck c o l l i d e d with the bus.' ' On the other hand, sentences l i k e (61) which do not take a symmetric predicate can have no more than two readings l i k e sentence (4), even i f they have the same kind of str u c t u r e as sentence (5). (61) Taroo to Hanako to Yosiko ga dekake-ta. go out a. 'Taro, Hanako, and Yoshiko went out (together).* b. 'Taro, Hanako, and Yoshiko a l l went out.' Sentence (6l/a) can be paraphrased as (62) Taroo to Hanako to Yosiko ga issyo n i dekake-ta. 'Taro, Hanako, and Yoshiko went out together.* 18 According to these f a c t s , a sentence which takes conjoined NP's as subject can have three readings when i t takes a symmetric predicate; on the other hand, such a sentence can have no more than two readings when i t does not take such a predicate. Next, compare sentences (5/a,) and (6l/,a) with the f o l l o w i n g p a i r s of sentences, i n which one,or two of the conjoined NP's i n (5/a) and ( 6 l/a) are moved out of subject p o s i t i o n . (63) a. Basu to torakku ga t a k u s i i to s y o o t o t u - s i - t a . 'The bus and the truck c o l l i d e d with the t a x i . 1 b. Basu ga torakku to t a k u s i i to s y o o t o t u - s i - t a . •The bus c o l l i d e d with the truck and the t a x i . ' (64) a. Taroo to Hanako ga Yosiko to dekake-ta. 'Taro and Hanako went out with Yoshiko.' b. Taroo ga Hanako to Yosiko to dekake-ta. 'Taro went out with Hanako and Yoshiko. 1 Semantically, sentences with and without a symmetric predicate behave i n d i f f e r e n t waysw . Sentence (63a) i s noncommittal as to whether or not the bus and the truck c o l l i d e d with each other, and sentence (63b) i s a l s o noncommittal as to whether or not the truck and the t a x i c o l l i d e d with each other. That i s , they are n e i t h e r synonymous with each other, nor with sentence ( 5/a). On the other hand, sentences (64a) and (64b) are synonymous not only with each other, but also with sentence ( 6 l / a ) . They can be paraphrased as (65a) and (65b) r e s p e c t i v e l y . (65) a. Taroo to Hanako ga Yosiko to i s s y o n i dekake-ta. 'Taro and Hanako went out together with Yoshiko.' 19 b. Taroo ga Hanako to Yosiko to issyo n i dekake-ta. •Taro went out together with Hanako and Yoshiko.' 7 1.2. There are some analyses dealing with the data presented so f a r . I w i l l examine them c l o s e l y i n the sections below. 1.2.1. Among such works, Okutsu (1967) was the f i r s t one done w i t h i n the framework of transformational grammar, although i t only dealt with sentences containing symmetric predicates. The f o l l o w i n g claims are made on the basis of the paraphrase r e l a t i o n s between such sentences. (66) ( i ) Sentences l i k e (3/a) are derived from sentences l i k e (1/a) by a transformational r u l e which i s formulated i n (67). ( i i ) Sentences l i k e (3/b) are derived from sentences l i k e (1/b) or (15) by a transformational r u l e which i s formulated i n (68). Note that the underlying s t r u c t u r e of sentence (1/b) i s considered to be the same as that of sentence (15). ( i i i ) Sentences l i k e (3/c) are derived from conjoined sentences l i k e (69) by Conjunction Reduction, though t h i s term i s not employed. (67) * 9 X [HPX C ] g ^ Y [ > 2 C 1 S y m m Z X [NP X Conj NP 2 Conj C ] S u b j Y Z (Subj»Subject, Symm=Symmetric, C«Case marker, Conj=Con-j u n c t i o n , OPT=Optional) OPT (68) X [NP X C Y [NP 2 C ] C o n c o m Z = > X [NP 1 Conj NP 2 Conj C ] S u b ^ Y Z (Concom=Concomi tat i ve) (69) Taroo ga kekkon-si, Hanako ga kekkon-si-ta. 2G 'Taro got married, and Hanako got married.' The weakness of Okutsu's claims l i e s i n the f a c t that there are a number of sentences which cannot be accounted f o r by the ru l e s of (67) and (68). Consider the f o l l o w i n g sentences: (70) Nihon ga Doitu to Amerika to t a t a k a t - t a . Japan Germany America f i g h t 'Japan fought Germany and America.' (71) Nihon ga Doitu to / Amerika to t a t a k a t - t a . 'Japan fought America with Germany.' Notice that these sentences take a symmetric predicate tatakat 'to f i g h t ' . Sentences (70) and (71) are both ambiguous, and they can convey the f o l l o w i n g readings r e s p e c t i v e l y . (70) Nihon ga Doitu to Amerika to t a t a k a t - t a . a. 'Japan fought Germany and America (at the same time). 1 - b. 'Japan fought Germany and America (separately).' c. 'Japan fought (some country) with Germany and America.' (71) Nihon ga Doitu to / Amerika to t a t a k a t - t a . a. 'Japan fought Germany with America.' b. 'Japan fought America with Germany.' As sentence (70/a) meets the SD ( S t r u c t u r a l Description) of the r u l e (67), t h i s r u l e can apply to y i e l d (72), the meaning of which i s e n t i r e l y d i f f e r e n t from that of (70/a). (72) Nihon to Doitu to Amerika ga t a t a k a t - t a . 'Japan, Germany, and America fought.' The second claim stated i n ,(66) implies that sentence (72) can be 21 synonymous with sentence (70/a), but t h i s i s not the case. Though sentence (72) has various readings, i t can never be synonymous with sentence (70/a). (72) Nihon to Doitu to Amerika ga t a t a k a t - t a . a. „ 'Japan, Germany, and America fought (one another).' b. 'Japan, Germany, and America fought (some country together).\ x c. 'Japan, Germany, and America fought (some country at d i f f e r e n t times).' Sentence (72) i s derivable not only from sentence (70/a), but also from sentences (70/c), (71/a) and (71/b). In the case of (70/c) rul e (68) i s a p p l i e d , and i n the case of (71/a) and (71/b) r u l e s (67) and (68) are both applied. However, sentence (72) can be synony-mous with n e i t h e r those sentences nor sentence (70/a). Thus, sentence (72) cannot be generated by Okutsu's grammar. The same kind of problem occurs i n the d e r i v a t i o n of sentence (73) i n the sense of (a) reading. (73) Taroo to Hanako ga kekkon-si-tagat-ta. want a. 'Taro and Hanako wanted to get married (to each other).' b. 'Taro and Hanako wanted to get married (to someone together).' c. 'Taro and Hanako (both) wanted to get married (to someone).' According to Okutsu, sentence (73/a) would have to be derived from the sentence below. (74) Taroo ga Hanako to kekkon-si-tagat-ta. 'Taro wanted to get married to Hanako.' 22 However, sentence (74) cannot be synonymous with sentence (73/a). Sentence (74) represents only Taro's i n t e n t i o n of marrying, while sentence (73/a) represents Taro's and Hanako's r e c i p r o c a l d e s i r e . Therefore, sentence (73/a) cannot be generated by Okutsu's grammar e i t h e r . 1.2.2. The proposals made by Kuno (1967-68) are mainly based on Lakoff and Peters (1966). One proposal states that a grammar of Japanese should contain a r u l e schema of the form (75) NP —» NP n t o , n>2 where (75) represents an i n f i n i t e set of r u l e s , each having the form NP NP • • • NP jto. These r u l e s generate base structures of the form (76) 1 0 NP NP • • • NP to Consequently, sentences l i k e (3/a) are generated by one of the above r u l e s ; on the other hand, to r e l a t e these sentences to sentences l i k e (1/a), Kuno proposes a r u l e of Conjunct Movement. This r u l e can be w r i t t e n as f o l l o w s : (77) Conjunct Movement SD: [[NP to] - NP] N p - VP OPT 1 2 3 > SC: 0 2 1+3* 1 I f we f o l l o w Kuno, sentence (1/a) would be derived from sentence (3/a) by Conjunct Movement. 23 The problems inherent i n Gkutsu's a n a l y s i s which were mentioned above can be solved using Kuno's proposals. Take the second problem as an example. The f o l l o w i n g would be the underlying structures of sentences (73/a) and (74) r e s p e c t i v e l y . (78) NP NP I Taroo Hanako NP, VP t a g a t 1 2 - t a NP NP to kekkon-si I I Taroo Hanako (79) Taroo NP NP VP tagat-ta I NP to kekkon-si Taroo Hanako By the a p p l i c a t i o n of the p r i n c i p l e and the r u l e stated i n footnote 10, structures (78) and (79) are converted i n t o (80) and (81) r e s p e c t i v e l y . / 24 (80) NP / \ NP to Hanako NP, I Taroo VP taga t - t a (81) NP NP kekkon-si / \ i NP to Hanako Taroo NP VP tagat-ta NP NP kekkon-si NP to Hanako * I Taroo To (80) Equi NP Deletion applies on the S Q cycle ;to delete NP 2, and then Casei^Marking applies to y i e l d the surface s t r u c t u r e of sentence (73/a). To derive the surface structure of sentence (74) from (81), Case.Marking applies f i r s t on the S^^ cycle and then Conjunct Movement stated i n (77) applies to y i e l d [ Taroo [ q Taroo ga Hanako to kekkon-b0 b l s i ] t a g a t - t a j . Next,' Equi NP Deletion applies on the S Q cycle before the a p p l i c a t i o n of Case Marking to delete the second Taroo i n S^. \ Let us now examine how sentences , (3/b) and (3/c) are derived. Kuno p o s i t s (82) and (83) as t h e i r respective underlying s t r u c t u r e s . (82) [dareka Taroo to J N p kekkon-si-ta,^ + jdareka Hanako to J kekkonr-si^ta. NP 25 (83) [dareka Taroo t o ] N p kekkon-si-ta + [dareka Hanako to ] N p kekkpn-si-ta These structures are d i f f e r e n t only i n that the predicate of each conjunct has an i d e n t i c a l index i n (82), while i t does not i n (83). The d e r i v a t i o n of sentence (3/b) from (82) i s as f o l l o w s : (84) [dareka Taroo to 3 H r kekkon-si-ta^ + ^dareka Hanako t o j NP kekkon-si-1& ± (*» (82.)) =^ [dareka to Taroo 3 N p ga kekkon-si-taj^ + [dareka to Hanako ~\ N p ga kekkon-si-ta^ Taroo ga dareka to kekkon-si-ta^ + Hanako ga dareka to kekkon-rsi-ta^ =^ Taroo ga kekkon-si-ta^ + Hanako ga kekkon-si-ta^ =^ [Taroo to Hanako ~\ N p ga kekkon-sir- t a ^ A f t e r the a p p l i c a t i o n of the u n i v e r s a l p r i n c i p l e and c e r t a i n r u l e s , Conjunct Movement, I n d e f i n i t e Pronoun Deletion and Conjunction Reduction apply i n that order. Sentence (3/c) i s derived from (83) i n the same manner. Thus i t would be safe to say that the d e r i v a t i o n of sentence (4/a) i s that shown i n (85). (85) Taroo ga gakkoo ,ve i t - t a ^ + Hanako ga gakkoo e i t - t a ^ —-^> [Tarpo, to Hanako ~\ N p ga gakkoo e i t - t a . ^ Then, how does Kuno r e l a t e sentence (3/b) to sentence (1/b)? He assumes that a grammar has a u n i v e r s a l p r i n c i p l e which replaces each conjunct with the other i n conjoined HP's. I f the p r i n c i p l e applies to the surface structure of (3/b), the structure w i l l be converted 26 i n t o [Hanako to Taroo] N p ga kekkon-si-ta^. By applying Conjunct Movement to t h i s s t r u c t u r e , the surface structure of (1/b) i s derived. Kuno's proposals presented above involve a number of serious problems. One of these i s r e l a t e d to h i s mechanism f o r generating a phrasal conjunction. As the mechanism i s too powerful, i t can produce underlying structures such, as (86 ) [ [ [Taroo Hanako to ][ N p Ziroo to ] N p Qzyon M e a r i i to ] K p t o ] N p kekkon-si-ta However, we do not have any utterance which corresponds to t h i s s t r u c t u r e . The second problem concerns the d e r i v a t i o n of the f o l l o w i n g sentence. (87) Kinoo kekkon-si-ta Taroo to Hanako wa UBC no gakusei da. yesterday 'Taro and Hanako who got married yesterday are students of UBC.' According to Kuno, state predicates (such as gakusei da 'to be a student*) can take n e i t h e r a phrasal conjunction nor an i d e n t i c a l index. Thus, i f i t i s the case that Taro and Hanako got married to each other, the underlying structure of sentence (87) would have to be something l i k e t h i s : 27 (88) 13 NP UBC no | gakusei da VP Taroo NP NP UBC no | gakusei da VP Hanako NP NP kinoo kekkon-L-ta NP A l s i -NP to Taroo NP kinoo kekkon-s i - t a NP to Hanako Hanako Taroo On the and c y c l e s , Case Marking applies f i r s t and then Conjunct Movement applies to y i e l d (89) as the subtree S^. (89) NP UBC no gakusei da I Taroo NP ga NP kinoo kekkon-s i - t a Taroo NP to Hanako As Conjunct Movement i s an o p t i o n a l r u l e , we do not have to apply i t ; however, i f we do not, we cannot get to the S Q c y c l e , f o r R e l a t i v i z a -14 t i p n i s blocked on the cycle because of a general c o n s t r a i n t . The structure (89) meets the SD of R e l a t i v i z a t i o n , and t h i s r u l e applies to y i e l d : 28 (90) L g [_ N p [_ s Hanako to kinoo kekkon-si-ta^ Taroo J UBC no gakusei da] The subtree becomes (91) i n the same manner. (91) [g [ N p Q s Taroo to kinoo kekkon-si-ta] Hanako] UBC no gakusei da] Now we must conjoin the two NP's NP^ and NP 2 i n order to get the surface s t r u c t u r e of (87). However, we do not have a r u l e to do i t . Although we have Conjunction Reduction, t h i s r u l e j u s t converts SQ i n t o (92). Note that Case Marking applies on the and S 2 c y c l e s . (92) Ts L^p L~NP L s Hanako to kinoo kekkon-si-ta3 Taroo] to ^NP ^-S T a r 0 ° *° k i n o ° kekkon-si-taj Hanako] ] ga UBC no 2 4 gakusei da] Therefore, I conclude that sentence (87) i n the sense of "those who got married to each other"cannot be generated by Kuno's grammar. The t h i r d problem i s r e l a t e d to Conjunct Movement. Kuno states that sentences(1/b) and (15) are derived from sentence (3/b) by t h i s r u l e , as sentence (1/a) i s from sentence (3/a). In the d e r i v a t i o n of.sentence (15), issyo n i i s attached to the r i g h t of the term which i s moved to the l e f t of ,the VP by Conjunct Movement, while i t i s not i n the de r i v a t i o n s of sentences (1/a) and (1/b)-,.,, To exp l a i n t h i s discrepancy, Kuno annexes the f o l l o w i n g conditions to Conjunct Movement. (93) ( i ) Issyo n i i s attached o p t i o n a l l y to the moved term, when 29 t h i s term i s a conjunct of conjoined NP's which are derived from conjoined sentences with i d e n t i c a l i n d i c e s , ( i i ) Issyo n i i s attached o b l i g a t o r i l y to the moved term,/when t h i s term i s a conjunct of doubly-conjoined NP's which are generated i n the base component, ( i i i ) Issyo n i must not be attaqhed to the moved term, when t h i s term i s a conjunct of conjoined NP's (but not doubly-conjoined ones) generated i n the base component. The existence of issyo n i i n sentence (15) i s accounted f o r by con-d i t i o n ( 9 3 i ) , because the sentence i s derived from the stru c t u r e underlying sentence (3/b), whose conjoined NP's are derived from conjoined sentences as shown i n (84). Condition ( 9 3 i i ) explains issyo n i i n sentence (94), the d e r i v a t i o n of which i s shown i n (95). (94) Mearii ga Zyon to issyo n i kenka-si-ta. (=Kuno's (6.27)) Mary John quarrel •Mary qua r r e l l e d (with someone) together with John.' (95) [ Npdareka t o j ^ N p Z y o n to M e a r i i ] 3 g a kenka-si-ta —^ > ^pZyon to Mearii~] ga dareka to kenka-si-ta Mearii.ga Zyon to issyo n i dareka to kenka-si-ta =^ Mearii ga Zyon to issyo n i kenka-si-ta (=(94)) Condition ( 9 3 i i i ) i s necessary to block the d e r i v a t i o n of sentence (15) from sentence (1/a), f o r these sentences cannot be synonymous. The above con d i t i o n s , however, are c l e a r l y not s u f f i c i e n t to account f o r the behavior of issyo n i . F i r s t , they cannot e x p l a i n why sentence (29) can take i s s y o n i . The sentence i s considered to be derived from the structure underlying sentence (3/b), the d e r i v a t i o n of 30 which i s shown in (84). Though Conjunct Movement applies in (84), issyo ni must not be attached to the moved term because of condition ( 9 3 i i i ) . Therefore, the phrase in (29) cannot be generated by Kuno's , grammar. Secondly, the conditions stated in (93) depend on whether the moved term is a conjunct of conjoined NP's derived from conjoined sentences or not. However, there is no way of knowing this. 1.2.3. Concerning sentence (3), Inoue (1976b) makes the following proposals: (9§) (i) Sentence (3/a) is derived from sentence (1/a) by a transformation which inserts a symmetric NP into a subject NP,. The derivation of sentence (3/a) is shown in (97a). (ii)<Sentence (3/b) is generated in the base component, and sentence (1/b) as well as sentence (15) is derived from i t by Conjunct Movement. The derivation of (1/b) and (15) is shown in (97b). ( i i i ) Sentence (3/c) is derived from conjoined sentences like (69) through Conjunction Reduction. The derivation of sentence (3/c) is shown in (97c). (97) a. Taroo ga Hanako to kekkon-si-ta. (=(l/a)) 'Taro.got married to Hanako." z=p Taroo to Hanako ga kekkon-si-ta. (=(3/a)) 'Taro and Hanako got married (to each other).' b. Taroo to Hanako ga kekkon-si-ta. (=(3/b) 'Taro and Hanako got married (to someone together).' ITaroo ga Hanako to kekkon-si-ta. (=(l/b)) 'Taro got married (to someone) with Hanako.' Taroo ga Hanako to issyo ni kekkon-si-ta. (=(15)) I'Taro got married (to someone) together with Hanako.'J c. Taroo ga kekkon-si, Hanako ga kekkon-si-ta. (=(69)) 'Taro got married, and Hanako got married.* Taroo to Hanako ga kekkon-si-ta. (=(3/c)) 'Taro and Hanako got married (to someone separately).' Therefore, sentence (4/a) would be generated i n the base component, and sentence (4/b) would be derived from conjoined sentences. Inoue's proposals involve a serious problem with r e l a t i o n to the derivation of sentences with a symmetric predicate. According to the proposal (96i), sentence (5/a) would have to be derived from either sentence (63a) or sentence (63b) by the insertion rule. However, on the basis of the fact that (5/a) can be synonymous with neither (63a) nor (63b), she claims that i t should be derived from sentence (98) through a deletion rule. (98) Basu to torakku to takusii ga basu to torakku to takusii to syoototu-si-ta. 'The bus, the truck, and the taxi c o l l i d e d with the bus, the truck, and the t a x i . ' This claim implies the p o s s i b i l i t y of sentence (3/a) being derived from the sentence given below. (99) Taroo to Hanako ga Taroo to Hanako to kekkon-si-ta. 'Taro and Hanako got married to Taro and Hanako.' However, she would deny the po s s i b i l i t y - o n the grounds of the accept-32 a b i l i t y difference between the following sentences: (100) Taroo to Hanako ga tagai n i kekkon-si-ta. each other •Taro and Hanako got married to each other.' (101) Basu to torakku to takusii ga tagai n i syoototu-si-ta. 'The bus, the truck, and the taxi c o l l i d e d with one another.' Her argument would be something l i k e t h i s : (i) as sentence (5/a) i s derived from sentence (93) via (101), i f sentence (3/a) i s derived from sentence (99) , (3/a) must be derived from (99) v i a the ungram-matical sentence (100); ( i i ) i t i s incorrect to derive a grammatical sentence from an ungrammatical one; ( i i i ) therefore, sentence (3/a) should not be derived from sentence (99). However, although I accept step ( i ) I do not accept ( i i ) , therefore I hesitate to accept the conclusion ( i i i ) . In addition, i f Inoue*s argument i s correct, sentence (102) w i l l never be derived from sentence (103), because sentence (104) i s not acceptable. (102) Taroo to Ziroo to Saburoo ga kyoodai da. brother •Taro, J i r o , and Saburo are brothers.* (103) Taroo to Ziroo to Saburoo ga Taroo to Ziroo to Saburoo to kyoodai da. 'Taro, J i r o , and Saburo are brothers of Taro, Ji r o and Saburo.' (104) XTaroo to Ziroo to Saburoo ga tagai n i kyoodai da. That i s , sentence (102) cannot be generated by Inoue's grammar. Note that kyoodai da 'to be a brother' i s a symmetric predicate, because i f (105a) i s true (105b) i s also true. 33 (105) a. Taroo ga Ziroo to kyoodai da. 'Taro is a brother of Jiro.' b. Ziroo ga Taroo to kyoodai da. 'Jiro is a brother of Taro.' To conclude, Inoue's grammar is inadequate in that i t permits the derivation of sentence (3/a) from not only sentence (1/a) but also from sentence (99). Inoue's other problem is related to the rule of Conjunct Movement. She does not give the formulation of the rule, but i t is not di f f i c u l t to suppose that she regards i t something like the one proposed by Kuno (i.e., (77)). If this is the case, Conjunct Movement would have to apply to sentences like (3/a) and (3/c) as well as to sentences like (3/b) , because they a l l have the same structure at the stage where the rule applies. If we apply the rule to sentences (3/a) and (3/c), we will get sentence (1) (i.e., Taroo ga Hanako to kekkon- si-ta) in both cases. However, this resulting sentence poses a problem. That i s , as sentence (3/a) is derived from sentence (1/a) as shown in (97a), the sentence which is derived from (3/a) by Con-junct Movement and the sentence from which (3/a) is derived become the same sentence; moreover, sentence (3/c) cannot be synonymous with the sentence which is derived from i t by the rule. This fact suggests that Conjunct Movement should not apply to sentences (3/a) and (3/c). That i s , Inoue's Conjunct Movement also needs some con-ditions as Kuno's does. My discussion of Kuno's conditions stated in (93) also applies to Inoue's. As sentences (3/a-c) have the same structure at the point when Conjunct Movement applies, i t is di f f i c u l t to prevent the rule from applying to sentences (3/a) and (3/c). Inoue might argue that this difficulty is overcome by arranging the rules in question in the proper order. To make Conjunct Movement applicable only to sentence (3/b), she might propose that the rule applies before Conjunction Reduction and the insertion rule which derives sentence (3/a) from (1/a). However, even i f the rules are arranged in this way, there is s t i l l a problem. In this case, the problem lies in the fact that the sentence derived from (3/b) by Conjunct Movement and sentence (1/a) are of exactly the same structure. Thus, i f the order is adopted, sentence (3/a) will be derived from the two sentences whose structures are the same but whose meanings are different. 35 Chapter 2 DERIVED CONJUNCTION 2.1. In the previous s e c t i o n s , I have reviewed the proposals of Okutsu (1967), Kuno (1967-68) and Inoue (1976b) which r e l a t e to the d e r i v a t i o n of sentences ( l ) - ( 5 ) , i n p a r t i c u l a r (1) and (3). Among these s t u d i e s , the l a t t e r two seem to present o b s e r v a t i o n a l l y adequate grammars. However, both of these grammars are not d e s c r i p t i v e l y adequate, because they do not always provide correct p r e d i c t i o n s about the r e l a t i o n s between the sentences. Their defect seems to stem from the f a c t that they contain the r u l e schema stated i n (75). The i n t r o d u c t i o n of the r u l e schema not only makes those grammars too powerful, but i n Kuno makes the formulation of Conjunct Movement more complicated. In Inoue, i t derives an ungrammatical sentence from two d i f f e r e n t underlying s t r u c t u r e s . Therefore, r e j e c t i n g the r u l e schema (75) I take the p o s i t i o n that sentences with conjoined NP's (such as sentences (3)-(5)) are a l l transformationally d e r i v a b l e . 2.2. Following Kuno and others, I assume that sentences l i k e (3/c) are derived from conjoined sentences by a r u l e of Conjunction Reduction. (3/c) Taroo to Hanako ga kekkon-si-ta. •Taro and Hanako got married (to someone s e p a r a t e l y ) . ' 2.2.1. Conjunction Reduction can be stated roughly as i n (106). 36 (106) Conjunction Reduction a. SD: [ [ X - A ] " - AND] 15 OPT 1 2 3 SC: [ l 0 b. SD: £[A - x]£ 1 2 SC: 1# [0 2 16 AND] 3 31B OPT B Condition: a l l occurrence of A are i d e n t i c a l . The above i s a Japanese version of Ross's formulation of Conjunct Movement. Therefore, the notation employed i n (106) i s exact l y what he uses i n h i s formulation of the r u l e . He s t a t e s : This notation should be in t e r p r e t e d to mean that i n any coordinate node of the category B, which dominates any number of conjuncts which are also of the category B, and each of which e i t h e r ends or begins with a constituent of category A, where a l l occurrences of A are i d e n t i c a l , a l l of these occurrences of A are superimposed, and adjoined to the conjoined node B. £Ross (1967): 220] Thus, Conjunct Movement derives sentence (3/c) from the underlying structure (107) i n the f o l l o w i n g manner: ^ (107) 17 Taroo ga dareka to Hanako ga dareka to 1 Q kekkon-si-ta kekkon-si-ta Conjunction Reduction 37 (108) S 2 AND dareka to | kekkon-si-ta NP, NP, Taroo ga Hanako ga ^ S r > S 2 Pruning (109) Taroo ga Hanako ga kekkon-si-ta Node Relabeling (110) NP, /X A NP. ga NP. ga I I 4 Taroo Hanako NP2 AND dareka to kekkon-si-ta >|j/ Conj miction Reduction (111) dareka to. kekkon-si-ta NP. NP„ AND I 1 I NP_ NP, 3 i 4 Taroo Hanako NP Pruning 38 (112) Taroo Hanako (113) NP, AND Hanako I1 Taroo S dareka to kekkon-si-ta VP dareka to kekkon-si-ta The nodes S 1 and S 2 in (108), which is generated by Conjunction Reduction, are pruned by a general convention. In (109) the node SQ is changed into NP by a special node relabeling convention which converts structures of the form £x X ANDJ^ into those of the form £x X AND]X, whenever Y immediately dominates conjoined structures of X. Then, Conjunction Reduction applies to NPQ in (110) to yield (111). After the general convention prunes NP^  and NP^  in (111), the universal principle and the rule stated in footnote 15 apply to (112) to yield 19 (113). Finally, Indefinite Pronoun Deletion applies to this structure to yield the surface structure of sentence (3/c). Sentence (30) is derived from the surface structure of (3/c) by an optional transfor-mation which adjoins the element tomo ni. (30) Taroo to Hanako ga tomo ni kekkon-si-ta. 39 •Taro and Hanako both got married (to someone).' 2.2.2. As I mentioned i n sec. 1.2.3, Inoue als o derives sentences l i k e (3/c) by Conjunction Reduction. However, the operation of her Conjunction Reduction i s e n t i r e l y d i f f e r e n t from that of mine. She regards i t as d e l e t i o n but I regard i t as adjunction. Therefore, i f we f o l l o w her, sentence (114) and (115) would be derived from the underlying structures (116) and (117) r e s p e c t i v e l y . (114) Taroo ga hon to z a s s i o k a t - t a . book magazine buy •Taro bought a book and a magazine;' (115) Taroo ga hon o kat-te yon-da. read * Taro bought and read the book.' Taroo ga hon o k a t - t a Taroo ga z a s s i o k a t - t a (Pred=Predicate) (117) S Taroo ga hon o k a t - t a Taroo ga hon o yon-da In the d e r i v a t i o n of (114), the c i r c l e d Pred and NP i n (116) are deleted by Conjunction Reduction; s i m i l a r l y , i n that of (115) the c i r c l e d NP's i n (117) are deleted. I f the boxed NP i n (117) 40 i s deleted by the r u l e , we w i l l get an ungrammatical sentence. (118) Taroo ga k a t ^ t e , hon o yon-da. Therefore, Inoue's r u l e of Conjunction Reduction must contain some co n d i t i o n l i k e (119) to block the d e r i v a t i o n of (118) from (117). (119) I f the i d e n t i c a l elements are n e i t h e r the leftmost nor the rightmost element of each conjunct, delete a l l i d e n t i c a l ele-ments except f o r the one of the leftmost conjunct. However, unfortunately t h i s c o n d i t i o n does not work w e l l . Consider the f o l l o w i n g structure:^ (120) NP < f l £ ) Pred Taroo ga Hanako n i hon o age-ta Taroo ga Ziroo n i hon o age-ta According to (119), the c i r c l e d NP hon o, as w e l l as the other c i r c l e d elements i n (120), should be deleted by Conjunction Reduction, but i f t h i s i s done an ungrammatical sentence w i l l r e s u l t . (121) Taroo ga Hanako n i hon o Ziroo n i age-ta. To get a grammatical sentence from (120), we must delete the boxed NP by Conjunction Reduction, not the c i r c l e d one. (122) Taroo ga Hanako to Ziroo n i hon. o age-ta. give * Taro gave a book to Hanako and J i r o . • 41 This fact seems to provide a piece of evidence against Inoue's Conjunction Reduction. On the other hand, Conjunction Reduction stated in (106) predicts that sentences (118) and (121) are ungrammatical. That i s , those sentences cannot be derived by the rule from the structures (123) and (124) which underlie sentences (115) and (122) respectively. (123) AND Taroo ga NP V Taroo ga NP V hon o kat-ta hon o yon-da (124) Hanako ni hon o age-ta Ziroo ni hon o age-ta Sentence (115) is derived from (123) via (125a) and (125b). (125) a. [ s [ N pTaroo g a ] [ y p [ y phon o kat-ta][ y phon o yon-daj AND]] b- CsUpTaro° s<HVp[NPhon 0 ] [ v [ v k a t - t a ] t v y o n - d a ] ^1]] Sentence (122) is derived from (124) via (126a-c). (126) a. [ s [ N p T a r o o g a ] [ y p [ypHanako ni hon o age-ta][ y pZiroo ni hon o age-ta] AND] ] 42 °- [ s C N P T a r o o g a ] [ v p [ N p [NpHanako ni hon o][ N pZiroo ni hon o] AND][vage-ta] ] ] c- [ S [ N P T a r 0 ° g a ] [ V p [ N p t N P H a n a k 0 n i ] [ N P Z i r 0 ° n i ] ^ 1 [ N phon o][ vage-ta]]] The above facts indicate that the rule of Conjunction Reduction stated in (106) seems to be descriptively more adequate than the one proposed by Inoue. 2.2.3. It is not always the case that Conjunction Reduction can apply to conjoined structures whenever they meet its SD. Observe the following pairs of sentences: (127) a. Watakusi wa ano hon to kono hon o yon-da. •I read that book and this book.' b. Watakusi wa ano to kono hon o yon-da. (128) a. Watakusi wa omosiroi hon to tumaranai hon o yon-da. interesting boring 'I read an interesting book and a boring book.' b. Watakusi wa omosiroi to tumaranai hon o yon-da. (129) a. Watakusi wa kinoo kat-ta hon to ototoi kat-ta hon o yon-da. yesterday buy the day before yesterday 'I read the book which I bought yesterday and the book which I bought the day before yesterday.' x b. Watakusi wa kinoo kat-ta to ototoi kat-ta hon o yon-da. (130) a. Watakusi wa o-kome to o-imo o kat-ta. Hon rice sweet potato •I bought rice and sweet potatoes.' x 20 b. Watakusi wa o-kome to imo o kat-ta. (131) a. Watakusi wa Amerika e to Kanada e i t - t a . Canada 43 b. Watakusi wa Amerika to Kanada e i t - t a . 'I went to America and Canada.' The structures of the underlined parts i n (a) sentences of (127)-(130) are i n f o r m a l l y represented i n (132), and the underlined part i n (131a) has the structure (133). (132) (133) 21 NP AND NP P NP P Det=Determiner Adj=Adjective Hon=Honorific Marker P=Particle The ungrammaticality of the (b) sentences of (127)-(130) i l l u s t r a t e s that Conjunction Reduction should not apply to structures l i k e (132), while the grammaticality of sentence (131b) shows that the t r a n s f o r -mation must apply to structures l i k e (133). That i s , structures (134a) and (134b) should not be derived from (132) by Conjunction Reduction, but s t r u c t u r e (135) must be derived from (133). (134) a. b. Hon N N N AND iDet) (Det) AND I 1 ! H4 44 (135) P To describe these f a c t s , a c o n d i t i o n l i k e (136) must be incorporated i n t o Conjunction Reduction stated i n (106). (136) When the category B i s NP, Conjunction Reduction must not apply to the structures i n which the category A i s a main constituent of the NP or H o n o r i f i c Marker, and i t must apply to the structures i n which the category A i s P a r t i c l e . Needless to say, the main constituent of NP i s N i n stru c t u r e (132) and NP i n structure (133). 2.3. In the sections below, I w i l l re-analyze sentences with issyo n i and f i n a l l y I w i l l propose that those sentences are derived from con-joined sentences with i d e n t i c a l i n d i c e s . 2.3.1. Consider the f o l l o w i n g sentences: (137) Taroo to Hanako ga issyo n i kuruma o k a t - t a . car buy •Taro and Hanako bought a car together.' (138) Taroo to Hanako ga issyo n i syoosetu o k a i - t e - i - r u . novel write •Taro and Hanako are w r i t i n g a novel together.* These sentences are ambiguous. (137) a. 'Taro and Hanako j o i n t l y bought a car.* b. 'Taro and Hanako bought cars together.' (138) a. 'Taro and Hanako are j o i n t l y w r i t i n g a novel.* 45 b. 'Taro and Hanako are w r i t i n g novels together.' Speaking of sentence (137), (b) reading i n d i c a t e s Taro and Hanako bought cars r e s p e c t i v e l y and that t h e i r actions of buying cars were performed at the same time and place. Thus the di f f e r e n c e between the two readings depends on whether the car(s) which they bought i s ( a r e ) e x a c t l y the same or not. The same i s true of sentence (138). I t s meaning di f f e r e n c e depends on whether or not the novel Taro i s w r i t i n g i s s t r i c t l y i d e n t i c a l to the one Hanako i s w r i t i n g . These f a c t s suggest that sentence (31) can also have two readings. (31) Taroo to Hanako ga issyo n i gakkoo e i t - t a . a. 'Taro and Hanako went to school [.the same school] together.' b. 'Taro and Hanako went to school [ d i f f e r e n t schools] together.' On the other hand, the sentences given below are not ambiguous. (139) Taroo to Hanako ga issyo n i z i s a t u - s i - t a . s u i c i d e do 'Taro and Hanako committed s u i c i d e together.;' (140) Tomoko to Yosiko ga issyo n i s i s s i n - s i - t a . f a i n t 'Tomoko and Yoshiko f a i n t e d together.' Note that sentences (139) and (140) take a s e l f - c o n t r o l l a b l e verb and a n o n - s e l f - c o n t r o l l a b l e verb r e s p e c t i v e l y . Those sentences i n d i c a t e only that the a c t i o n expressed by a verb was performed by the conjoined NP's (e.g., Taro and Hanako i n (139)) at the same time and place. How does i t come about that sentences l i k e (137) have two readings while those l i k e (139) have only one reading? One reason that I can 46 think of at present i s that sentences l i k e (137) have an NP i n a d d i t i o n to a conjoined NP while those l i k e (139) have only a conjoined NP. Yet, t h i s l i n e of reasoning might be f a l l i b l e , since we have sentences which have an NP i n a d d i t i o n to a conjoined NP and which do not carry two readings. (141) Watakusi ga pen to enpitu o issyo n i k a t - t a . pen p e n c i l • I bought a pen and a p e n c i l together.' (142) Watakusi ga hon to tegami o issyo n i okut-ta. l e t t e r send 'I sent the book and the l e t t e r together.' In these sentences, the reason seems to l i e i n the f a c t that we cannot imagine two d i f f e r e n t watakusi's. Then, how should we deal with the sentences given below? (143) Watakusi wa Taroo to Hanako n i issyo n i sen-en-satu o age-ta. 1000 yen b i l l give ' ( L i t . ) I gave a 1000-yen b i l l to Taro and Hanako together.' (144) Watakusi wa kono mondai o seizi-men to keizai-men kara issyo problem p o l i t i c s aspect economy from n i kangae-te-mi-ta. think t r y ' ( L i t . ) I t r i e d to think over t h i s problem from a p o l i t i c a l aspect and an economical aspect together.' Sentence (143) can have two readings: i n one reading, Taro and Hanako as a u n i t got a 1000-yen b i l l , and i n the other reading they each got one. However, sentence (144) seems to have only one reading, because the NP kono mondai ' t h i s problem' r e f e r s to a s p e c i f i c t h i n g l i k e watakusi i n (141). Prom the data given so f a r , we could say that a sentence with issyo n i can have two readings, when i t has not only 47 a conjoined NP but a l s o a NP whose referent i s ambiguous. Next, l e t us consider the meaning of issyo n i . I t i n d i c a t e s temporal and s p a t i a l togetherness. Thus i t i s not possible f o r issyo n i to co-occur with a conjoined NP whose conjuncts i n d i c a t e place or time, because i t i s impossible to do something at d i f f e r e n t times i n the manner of "togetherness" and a l s o impossible to be i n or to go to d i f f e r e n t places i n the manner of "togetherness". This p r e d i c t i o n i s confirmed by the f a c t that sentences (145)-(149) are a l l unacceptable. (145) XTaroo ga n i - z i to s a n - z i n i issyo n i gakkoo e i t - t a . two o'clock three * xTaro went to school at two o'clock and three o'clock together.' (146) XTaroo ga 1970-nen to 1972-nen n i issyo n i z i s a t u - s i - t a . year ' xTaro committed s u i c i d e i n 1970 and 1972 together.' X (147) Hon-ya ga n i - k a i to san-kai n i issyo n i ar-u. bookstore 2nd f l o o r 3rd on e x i s t ' xThe bookstore i s on the second f l o o r and t h i r d f l o o r together.' X (148) Taroo ga gakkoo to kyookai e issyo n i i t - t a . church | XTaro went to school and church together.' (149) XTaroo ga kono m i t i to ano m i t i o issyo n i t o o t - t a . road go along , XTaro went along t h i s road and that road together.' To summarize: (150) ( i ) When a sentence with issyo n i contains not only a conjoined NP but also an NP whose referent i s ambiguous, the sentence can have two readings. I f the referent of the NP i s not ambiguous, the sentence receives only a " j o i n t i n t e r p r e t a t i o n " . 48 ( i i ) When NP's which i n d i c a t e place or time are conjoined, the conjoined NP's should not co-occur with i s s y o n i . 2.3.2. Compare sentences (137)-(142) with the f o l l o w i n g sentences, where one of the conjoined NP's i s moved to the p o s i t i o n before i s s y o  n i . (151) Taroo ga Hanako to issyo n i kuruma o k a t - t a . 'Taro bought a car together with Hanako.* (152) Taroo ga Hanako to issy o n i syoosetu o k a i - t e - i - r u . 'Taro i s w r i t i n g a novel together with Hanako.* (153) Taroo ga Hanako to issyo n i z i s a t u - s i - t a . 'Taro committed s u i c i d e together with Hanako.' (154) Tomoko ga Yosiko to issyo n i s i s s i n - s i - t a . 'Tomoko f a i n t e d together with Yoshiko.' (155) Watakusi ga pen o enpitu to issy o n i k a t - t a . 'I bought a pen together with a p e n c i l . ' (156) Watakusi ga hon o tegami to i s s y o n i okut-ta. ' I sent the book together with the l e t t e r . * These sentences are synonymous with sentences (137)-(142) r e s p e c t i v e l y , though sentences (151) and (152) are ambiguous and sentences (153)-(156) are not. However, i t i s not always the case that sentences which take conjoined NP's as objects are semantically equivalent to those where one of the conjoined NP's i s moved. (157) Taroo ga Tomoko to Yosiko o issyo n i mi-ta. look at •Taro looked at Tomoko and Yoshiko together.' (158) Taroo ga Tomoko o Yosiko to issyo n i mi-ta. 'Taro looked at Tomoko together with Yoshiko.' 49 Sentence (157) has only one reading, but sentence (158) can have two readings. (158) a. 'Taro looked at Tomoko and Yoshiko together.' b. 'Taro and Yoshiko looked at Tomoko together.' Sentences (158/a) and (158/b) are synonymous with sentences (157) and (159) r e s p e c t i v e l y . (159) Taroo to Yosiko ga issyo n i Tomoko o mi-ta. 'Taro and Yoshiko looked at Tomoko together.' Note1 that sentence (159) i s not ambiguous and that i t i s synonymous with sentence (160). (160) Taroo ga Yosiko to issyo n i Tomoko o mi-ta. 'Taro looked at Tomoko together with Yoshiko.' That i s , sentences (158/b), (159) and (160) are synonymous with one another. Next, compare sentence (143) with the f o l l o w i n g sentence: (161) Watakusi wa Taroo n i Hanako to issyo n i sen-en-satu o age-ta. •I gave a 1000-yen b i l l to Taro together with Hanako.' Sentence (161) i s five-way ambiguous: i t can carry not only the two readings sentence (143) has, but also the readings that sentences (162) and (163) have. (162) Hanako to watakusi wa issyo n i Taroo n i sen-en-satu o age-ta. a. 'Hanako and I j o i n t l y gave a 1000-yen b i l l to Taro.' b. 'Hanako and I each gave a 1000-yen b i l l to Taro.' 5G (163) Watakusi wa Taroo n i sen-en-satu to Hanako o is s y o n i age-ta. •I gave Taro a 1000-yen b i l l and Hanako (as a bride) together.' In (a) reading of (162) Taro gets only 1000 yen, but i n (b) reading he gets 2000 yen. Sentence (163) i s unusual, although i t i s i n t e r -pretable i n the sense that Hanako i s given as a b r i d e ; however, i f Hanako of (163) i s replaced by hon, the sentence i s quite normal as shown i n (164). (164) Watakusi wa Taroo n i sen-en-satu to hon o i s s y o n i age-ta. •I gave Taro a 1000-yen b i l l and a book together.' Note that sentences (163) and (164) are both unambiguous. Prom the f a c t s presented above, we could say the f o l l o w i n g : (165) Sentences i n which one of the conjoined NP's i s moved to the p o s i t i o n before issyo n i are synonymous with sentences with the conjoined NP's; moreover, they can have other readings. 2.3.3. As f o r the surface structures of sentences l i k e (137), we can propose two kinds of t r e e s . The d i f f e r e n c e between them i s due to the s t r u c t u r a l p o s i t i o n of i s s y o n i . Thus, we could propose e i t h e r (166a) or (166b) as the surface s t r u c t u r e of sentence (137). (166) a. NP ga i s s y o n i NP V / X A s I NP NP NP o k a t - t a / \ I I NP to Hanako kuruma I Taroo 51 b. S NP VP N P ' / ^ S S V g a ? NP V /\ A I NP NP issyo n i NP o k a t - t a A I \ NP to Hanako kuruma I Taroo Note that the node l a b e l dominating issyo n i i s represented with a question mark i n both s t r u c t u r e s . Considering the grammaticality of the pseudo-cleft sentences of (137), one might claim that the node should be NP. (167) a. Taroo to Hanako ga kuruma o k a t - t a no wa issy o ( n i ) da. ' ( L i t . ) The way Taro and Hanako bought a car i s together.' b. Issyo n i kuruma o k a t - t a no wa Taroo to Hanako da. •The persons, who bought a car together are Taro and Hanako.' c. Taroo to Hanako ga issyo n i k a t - t a no wa kuruma da. 'What Taro and Hanako bought together i s a car.' .. The element which i s moved before the copula da by Pseudo-cleft Sen-23 tence Formation i s u s u a l l y considered to be an NP. Therefore, the grammaticality of sentence (167a) seems to provide a piece of evidence i n support of the claim. However, observe the f o l l o w i n g sentence: (168) Hanako ga benkyoo-si-das-ita no wa kyuu n i da. study begin suddenly ' ( L i t . ) The way Hanako began to study i s suddenly.' This sentence i s acceptable as i s sentence (167a). Sentence (169) 52 underlies sentence (168). (169) Hanako ga kyuu n i benkyoo-si-das-ita. •Hanako began to study suddenly.' No one would claim that the phrase kyuu n i 'suddenly* i n sentence (169) i s an NP. This suggests that the element moved by Pseudo-cleft Sentence Formation i s not always an NP. Such adverbs as kyuu n i can also be moved by t h i s transformation. Thus the grammaticality of sentence (167a) cannot simply lead us to the conclusion that issyo n i i n sentence (137) i s an NP. In Japanese a l l NP's can be t o p i c a l i z e d at l e a s t i n a simple sentence. In a d d i t i o n , the t o p i c a l i z e d phrases are only NP's. The sentences i n (170) are derived from sentence (171) by T o p i c a l i z a t i o n . (170) a. Boku wa Hanako n i tosyokan de a t - t a . I l i b r a r y at meet •As f o r myself, I met Hanako at the l i b r a r y . ' b. Hanako n i wa boku ga tosyokan de a t - t a . •As f o r Hanako, I met her at the l i b r a r y . ' c. Tosyokan de wa boku ga Hanako n i a t - t a . 'As f o r at the l i b r a r y , I met Hanako there.' (171) Boku ga Hanako n i tosyokan de a t - t a . 'I met Hanako at the l i b r a r y . ' The grammaticality of sentences (170a)-(170c) shows that the phrases boku ga. Hanako n i and tosyokan de i n sentence (171) are a l l NP's. One might argue that kyuu n i i n sentence (169) i s also an NP,.sincev the phrase has the same fu n c t i o n ( i . e . , that of modifying predicates) as ordinary NP's (e.g., Hanako n i i n (171)). However, the argument 53 i s not c o r r e c t , because sentence (172), i n which kyuu n i i n sentence (169) i s t o p i c a l i z e d , -is not acceptable. (172) xKyuu n i wa Hanako ga benkyoo-si-das-ita. The u n a c c e p t a b i l i t y of t h i s sentence demonstrates that kyuu n i i s not an NP. 2^ As sentence (173), i n which iss y o n i i n sentence (137) i s t o p i c a l i z e d , i s also unacceptable, t h i s phrase i s not an NP e i t h e r . (173) X I s s y o n i wa Taroo to Hanako ga kuruma o k a t - t a . The f a c t s presented so f a r i l l u s t r a t e that kyuu n i and issyo n i are s y n t a c t i c a l l y very s i m i l a r . The phrase kyuu n i i s considered to be an adverb; therefore, i t would be safe to say that the node with a question mark i n (166a) and (166b) i s Adv ( i . e . , Adverb). Then, which of the two structures i s more appropriate as the surface structure of sentence (137)? I have shown i n footnote 17 that VP's can be replaced with soo-si 'to do so' by the replacement transformation. The phrase issyo n i kuruma o kat i n sentence (137) can be replaced by soo-si as i n (174). (174) Taroo to Hanako ga issy o n i kuruma o k a t - t a . Zyon to M e a r i i mo s o o - s i - t a . 'Taro.and Hanako bought a car together.' 'John and Mary d i d so too.' This f a c t i n d i c a t e s that issyo n i i s a part of the VP. Thus I conclude that sentence (137) has as i t s surface structure (166b') i n which the node ? i n (166b) i s replaced by Adv. 54 (166) b'. NP NP issyo n i NP o k a t - t a A I NP to Hanako kuruma i Taroo As a r e s u l t , sentences (141) and (143) are considered to have the f o l l o w i n g surface structures r e s p e c t i v e l y . (175) (176) pen to enpitu o issyo n i k a t - t a Taroo to Hanako n i issyo n i sen-en-satu o age-ta Next, l e t us consider the surface structures of sentences l i k e (151), which do not take conjoined NP's. Sentence (151) could have the f o l l o w i n g two p o s s i b i l i t i e s as i t s surface s t r u c t u r e . 55 (177) a. Hanako to issyo n i kuruma o ka t - t a (Adv) kuruma o k a t - t a Hanako to issyo n i I consider (177a) more appropriate as the surface st r u c t u r e of sentence (151) than (177b) f o r the f o l l o w i n g reason. Consider sentence (178). (178) Taroo ga Hanako to / kuruma o issyo n i k a t - t a . 'Taro bought a car together with Hanako.' To derive t h i s sentence from (177b), we must p o s i t a transformation which i n s e r t s an element i n t o an ad v e r b i a l phrase or moves an element of an adverbial phrase out of that phrase. Howeverthere are no independent reasons f o r such a transformation. On the other hand, to derive sentence (178) from (177a) we can make use of a r u l e of Scrambling, which can be w r i t t e n i n t h i s way: 25 (179) Scrambling jNP 1 " t A d v i ~ OPT 1 SC: 1 2 4 3 3 4 2 X 5 5 56 This transformation is also used to relate sentence (180) with sentence (180) Watakusi wa kinoo tosyokan kara hon o kari-ta. yesterday borrow 'I borrowed a book from the library-yesterday.' (181) Watakusi wa tosyokan kara kinoo hon o kari-ta. 'I borrowed a book from the library yesterday.' That i s , in the derivation of sentence (178) from (177a) we need not posit a new transformation. This indicates that (177a) is more appro-priate than (177b) as the surface structure of (151), for the former does not make the grammar more complicated than the latter does. 2.3.4. In sec. 2.1 I have rejected the claim that sentences with conjoined NP's are generated in the base component. Alternatively, I propose that sentences like (137) which contain such NP's are derived from conjoined sentences whose predicates take identical indices. Thus sentence (137) would be derived from the underlying structure given below: (181). (182) S NP Taroo kuruma kat-ta. kuruma kat-ta The symbol i indicates that the constituents with i t are identical. To get the surface structure of sentence (137) (i.e., (166b')), Case Marking applies f i r s t on the S 1 and S 2 cycles, 57 (183) AND NP, ga NP V I A l l' A I Taroo NP o kat-ta.. Hanako NP o k a t - t a —————x i • 1 kuruma kuruma As the r e s u l t i n g structure (183) meets the SD of Conjunction Reduction stated i n (106), the r u l e applies on the S Q cycle to y i e l d (184). (184) A .A NP, ga NP. ga ) 3 i 4 NP kuruma o ka.t-t&J Taroo Hanako Note that a V l i k e k a t - t a ^ i n (184) which i s derived by the a p p l i c a t i o n of Conjunction Reduction to V s with i d e n t i c a l i n d i c e s i s considered to be a marked element. The general convention and the s p e c i a l node r e l a b e l i n g convention stated i n sec. 2.2.1 apply to (184) to prune the nodes S^ and S 2 and to change the node SQ i n t o NP. 58 (185) NP. NP-NP„ AND • MP V A I NP, ga NP, ga 3 i 4 NP o k a t - t a . Taroo Hanako kuruma As NP Q i n (185) also meets the SD of Conjunction Reduction, the r u l e applies on the S cycle to y i e l d (186). (186) NP. NP_ AND I I NP- NP, I I Taroo Hanako The general convention applies to t h i s s t r u c t u r e again to prune the nodes NP^ and NP^. Next, Issyo n i I n s e r t i o n , the formulation of which i s something l i k e (187), applies to (186) without those NP's to y i e l d (188). 26 (187) Issyo n i I n s e r t i o n SD: X - [ N p C N p X AND] P j - X - V 1 ' 2 3 4 SC: 1 2 iss y o ni+3 Condition: 4 i s a marked element. OPT 59 (188) Taroo Hanako kuruma F i n a l l y , the general p r i n c i p l e and the r u l e stated i n footnote 15 apply to the node NP Q i n (188) to y i e l d (189). (189) issyo n i NP o k a t - t a NP AND Hanako kuruma Taroo The structure i n which AND i n (189) i s r e a l i z e d as to i s the surface structure of sentence (137) ( i . e . , (166b 1)). Similarly,'sentence (141) i s derived from the underlying structure given below: (190) AND NP NP VP watakusi NP V watakusi H I > v, pen kat-ta. enpitu k a t - t a ^ 60 I t should be noted that i f Issyo n i I n s e r t i o n does not apply to (186), (186) w i l l be converted i n t o (191) by the general p r i n c i p l e and other r u l e s . (191) / A A I NP NP NP o k a t - t a , / \ I i NP AND Hanako kuruma Taroo The s t r u c t u r e i n which AND i n (191) i s r e a l i z e d as to i s the surface s t r u c t u r e of the f o l l o w i n g sentence. (192) Taroo to Hanako ga kuruma o k a t - t a . •Taro and Hanako bought a car (together).' I have mentioned before that sentence (137) has the two readings (a) and (b). I assume that the meaning d i f f e r e n c e depends on whether or not the NP kuruma i n the f i r s t conjunct ( i . e . , S^) of (182) and the NP kuruma i n the second conjunct (Sg) are s t r i c t l y i d e n t i c a l . That i s , sentence (137/a) i s derived when the referents of those NP's are the same, and sentence (137/b) i s derived when they are not. I t i s obvious that t h i s claim i s not w i l d , when the f o l l o w i n g English sentence i s considered c a r e f u l l y . (193) John scratched h i s arm and so d i d Mary. 61 According to Ross (1967: 189), sentence (193) i s ambiguous—it could be derived from the st r u c t u r e underlying sentence (194) or the one underlying sentence (195). (194) John i scratched h i s i arm and Mary^ scratched ber^ arm too. (195) John^^ scratched h i s i arm and Mary^ scratched h i s i arm too. The problem i s that the differ e n c e between h i s arm and her arm i n sentence (194) i s disregarded i n the d e r i v a t i o n of sentence (193). In other words, though the NP's are not s t r i c t l y i d e n t i c a l , sentence (193) i s derived from the struc t u r e underlying (194). Thus Ross (1969: 268) claims that any theory i n which (193) can be derived from,(194) must contain a d e f i n i t i o n of "sloppy i d e n t i t y " , i n which i t i s s p e c i f i e d e x a c t l y what differences can be disregarded. This claim c l e a r l y corresponds to my assumption. Therefore, i f Ross's 27 claim i s accepted, mine should a l s o be accepted. Next, I w i l l e x p l a i n how sentences l i k e (151) are derived. Since these sentences are synonymous with sentences l i k e (137), I propose that they should be derived from the same underlying structures. Thus sentence (151) i s derived from the structure underlying sentence (137). I t s d e r i v a t i o n would be as f o l l o w s : (196) a. [„ \„ Taroo kuruma kat-ta,If,. Hanako kuruma k a t - t a "1 AND] L S0 S l 1 *2 (=(182)) ^ b' ts WUpb L u p ^ a r o o ^ H a n a k o ] AND] ga] issyo n i kuruma o kat-ta.1 (=(188)) V 62 NP„ NP, AND issy o n i NP o k a t - t a 2 i ± i J-Hanako Taroo kuruma d. NP NP A I NP 2 AND Taroo issyo n i NP o k a t - t a . I kuruma Hanako e. NP Adv V . . ga NP / \ -A I Taroo NP_ AND issyo n i NP o k a t - t a i — i 1 Hanako kuruma Needless to say, the structure (196b) i s derived from (196a) i n the same manner (188) i s derived from (182). To derive (196c) from (196b), an o p t i o n a l transformation applies to NP^ i n (196b) to change the order of the conjoined NP's ( i . e . , NP 1 and NPg). This transformation, which I c a l l NP Scrambling i n NP, can be formulated i n t h i s way: 63 (197) NP Scrambling i n NP SD: X - [ w p NP - X - NP - X AND] - X OPT 1 2 3 4 5 6 > SC: 1 4 3 2 5 6 To the s t r u c t u r e (196c), the general p r i n c i p l e and the r u l e stated i n footnote 15 apply to y i e l d (196d). The r u l e needed to derive (196e) from (196d) i s Conjunct Movement, the formulation of which would be that shown i n (198). (198) Conjunct Movement S D V x - [ N p U P N P A N D ^ - X P ] - X - V O P T 1 2 3 4 5 ••: ' > SC: 1 0 3 2+4 5 Condition: 5 i s a marked element. The structure i n which AND i n (196e) i s r e a l i z e d as to i s the surface stru c t u r e of sentence (151) ( i . e . , (177a)). The i m p l i c a t i o n of the proposal which I have presented i n t h i s s e c t i o n i s that concomitative NP's are not generated i n the base component. They are derivable only by Conjunct Movement stated i n (198). Therefore, i f t h i s r u l e a pplies c y c l i c a l l y i t can apply to a conjunct of conjoined sentences, i n which case sentence (199) would be derived from the two e n t i r e l y d i f f e r e n t underlying structures (200) and (201). (199) Taroo to Ziroo ga Hanako to gakkoo e i t - t a . 'Taro and J i r o went to school together with Hanako.' 64 (200) AND Taroo ga gakkoo e i t - t a . Ziroo ga gakkoo e it - t a , Hanako ga gakkoo e it - t a , (201) AND AND Taroo ga Hanako ga gakkoo e gakkoo e i t - t a ^ i t - t a ^ Ziroo ga Hanako ga gakkoo e gakkoo e i t - t a . i t - t a , The derivations of (199) from (200) and (201) would be roughly (202) and (203) respectively. (202) (200) jj^ Conjunction Reduction (on the S Q cycle) Taroo to Ziroo gakkoo e to Hanako ga . i t - t a . NP Scrambling i n NP and Conjunct Movement Taroo to Ziroo ga Hanako to gakkoo e i t - t a , 65 (203) (201) Conjunction Reduction (on the S- and S 2 cycles) AND Taroo to Hanako ga Ziroo to Hanako ga gakkoo e i t - t a i gakkoo e i t - t a ^ It NP Scrambling i n NP and Conjunct Movement (on the S & and cycles) AND Taroo ga Hanako to Ziroo ga Hanako to gakkoo e i t - t a ^ gakkoo e i t - t a ^ ^ Conjunction Reduction (on the S Q cycle) Taroo to Ziroo ga Hanako to gakkoo e i t - t a . 1 I t i s undesirable to derive an unambiguous sentence from more than two d i f f e r e n t underlying s t r u c t u r e s , so we must block one of the de r i v a t i o n s . I f we admit deriva t i o n s l i k e (203), we w i l l have several underlying structures f o r the f o l l o w i n g unambiguous sentence. (204) Taroo to Ziroo to Saburoo ga Hanako to issyo n i gakkoo e i t - t a . 'Taro, J i r o , and Saburo went to school .together with Hanako.' 66 This fact suggests that we must block the derivation of (199) from (201). To block i t we must prevent Conjunct Movement from applying cyclically. This means that the rule applies only on the last cycle. Needless to say, i t must apply after Conjunction Reduction. I have mentioned in sec. 1.1.3-2 that issyo ni should not co-occur with state predicates while tomo ni can. (26) a. Taroo wa Hanako to issyo ni UBC no gakusei da. b. Taroo wa Hanako to tomo ni UBC no gakusei da. •Taro and Hanako are both students of U B C (40) a. xSatoo-san to Suzuki-san wa issyo ni Ninon-zin da. b. Satoo-san to Suzuki-san wa tomo ni Nihon-zin da. •Mr. Sato and Mr. Suzuki are both Japanese.' The sentences in (26) and (40) are considered to be derived from con-joined sentences. For example, sentence (40b) is derived from the underlying structure (205) by Conjunction Reduction. (205) S S S AND NP VP NP VP Satoo-san ga Nihon-zin da Suzuki-san ga Nihon-zin da Thus, to block the derivations of (26a) and (40a), Issyo ni Insertion and Conjunct Movement must have another condition, which says: (206) V is not a state predicate. However, the grammaticality of sentence (26b) suggests that Conjunct 67 Movement applies even when V i s a state predicate. To account f o r t h i s f a c t , the conditions of the r u l e must be r e v i s e d i n t h i s way: (207) Conditions of Conjunct Movement: (1) 5 ( i . e . , V) i s a marked element and i s not a state predicate, or ( i i ) 4 contains tomo n i and 5 i s a state predicate. Condition ( i i ) p r edicts that sentence (52) i s not grammatical. (52) Taroo wa Hanako to UBC no gakusei da. The structure underlying sentence (52) does not contain tomo n i although i t has a state predicate, so i t does not meet cond i t i o n ( i i ) . Thus Conjunct Movement cannot apply to the struc t u r e to y i e l d (52). 2.4. In the sections below, I w i l l discuss how sentences with a symmetric predicate are derived. 2.4.1. I assume that sentences (3/b) and (70/a) have (208) and (209) as t h e i r respective underlying s t r u c t u r e s . (3/b) Taroo to Hanako ga kekkon-si-ta. •Taro and Hanako got married (to someone together).' (70/a) Nihon ga Doitu to Amerika to t a t a k a t - t a . •Japan fought Germany and America (at the same time).* (208) AND Taroo ga dareka to Hanako ga dareka to kekkon-si-ta. kekkon-si-ta. 68 (209) Doitu to tatakat-ta. Amerika to tatakat-ta^ Sentence (3/b) is derived from (208) in the same manner sentence (3/c) is derived from (107). Sentence (70/a) is derived from (209) by Conjunction Reduction via (210a) and (210b). (210) a. [_s L*NpNihon g a ] [ y p f y pDoitu to tatakat-tajf^Amer ika to tatakat-ta ±1 AND ] ] b « [ S C N P N i h o n S a ] [ V p C N p t N P D o i t u to][NpAn»erika to] AND] tatakat-ta^l3 Sentences (3/b) and (70/a) have (211) and (212) as their respective intermediate structures. (211) NP ga kekkon-si-ta^ NP NP AND Taroo Hanako 69 (212) S NP VP Nihon ga NP V / \ I NP to t a t a k a t - t a NP NP AND Doitu Amerika Notice that the predicates of these structures are marked elements. The structures (211) and (212) both meet the SD of Issyo n i i n s e r t i o n with revised conditions (206), so i f the r u l e applies we w i l l get sentences (213) and (214). (213) Taroo to Hanako ga issyo n i kekkon-si-ta. (=(29)) •Taro and Hanako got married (to someone) together.' X (214) Nihon ga Doitu to Amerika to issyo n i t a t a k a t - t a . Sentence (214) i s unacceptable i n the sense of sentence (70/a). In a d d i t i o n , i f we apply Conjunct Movement to the structures which are derived from (211) and (212) by applying NP Scrambling i n NP and some other r u l e s , we w i l l get the f o l l o w i n g sentences: (215) a. Taroo ga Hanako to kekkon-si-ta. (=(l/b)) •Taro got married (to someone) together with Hanako.' b. Taroo ga Hanako to issyo n i kekkon-si-ta. (=(15)) 'Taro got married (to someone) together with Hanako.' (216) a. xNihon ga Doitu to / Amerika to t a t a k a t - t a . b. XNihon ga Doitu to / Amerika to issyo n i t a t a k a t - t a . 70 Notice that not only Conjunct Movement but also Issyo n i I n s e r t i o n i s applied i n (b) sentences. The sentences i n (216) are unacceptable l i k e sentence (214) i n the sense of sentence (70/a). As the s t r u c t u r e (212) d i f f e r s from (-211) i n that symmetric NP's are conjoined, the un a c c e p t a b i l i t y of sentences (214) and (2l6a-b) suggests that Issyo n i In s e r t i o n and Conjunct Movement should not apply to the structures i n which symmetric NP's are conjoined. Therefore, i n order to prevent the transformations from applying to such s t r u c t u r e s , a c o n d i t i o n l i k e (217) i s necessary. (217) The conjoined NP's are not symmetric ones. This c o n d i t i o n blocks the deriva t i o n s of-sentences (214) and (2l6a-b) from (212). As a r e s u l t of (217), Issyo n i I n s e r t i o n and Conjunct Movement have (218) and (219) as t h e i r respective conditions. (218) Conditions of Issyo n i I n s e r t i o n : ( i ) and ( i i ) and ( i i i ) ( i ) 4 ( i . e . , V) i s a marked element, ( i i ) 4 i s not a state predicate. (=(206)) ( i i i ) The conjoined NP's are not symmetric ones. (=(217)) (219) Conditions of Conjunct Movement: ( ( i ) or ( i i ) ) and ( i i i ) ( i ) 5 ( i . e . , V) i s a marked element and i s not a state predicate. (=(207i)) ( i i ) 4 contains tomo n i and 5 i s a state predicate. (=(207ii)) ( i i i ) The conjoined NP's are not symmetric ones. (=(217)) 2.4.2. I assume that a Japanese grammar contains a r u l e of Reciprocal Transformation, which can be represented schematically i n (220). 71 (220) Reciprocal Transformation a. S S S S AND NP VP NP1 NP2 AND •••V. ar3 i ; 4 i b. NP NP S NP S NP 1 S NP2 I consider that this transformation can apply only when each conjunct meets the following conditions: (221) Conditions of Reciprocal Transformation: (i) and ( i i ) and ( i i i ) ( i ) Each conjunct takes the same symmetric predicate, ( i i ) Each conjunct i s id e n t i c a l except f o r a subject NP (or a head NP) and a symmetric NP. ( i i i ) A subject NP (or a head NP) and a symmetric NP of the f i r s t conjunct are i d e n t i c a l with a symmetric NP and a subject NP (or a head NP) of the second conjunct respectively. ( i . e . , NP^NP^, NP^NP^) 2.4.2.1. Reciprocal Transformation stated i n (220a) applies to the structure (222) to y i e l d sentence (3/a). 72 (222) Taroo ga NP V Hanako ga NP V Hanako to kekkon-si-ta, Taroo to kekkon-si-ta. (3/a) Taroo to Hanako ga kekkon-si-ta. •Taro and Hanako got married (to each other).' Reciprocal Transformation stated i n (220b) applies to the struc t u r e (223), which i s derived from (224), to y i e l d the struc t u r e underlying sentence (87) i n the sense of "those who got married to each other''. (87) Kinoo kekkon-si-ta Taroo to Hanako wa UBC no gakusei da. •Taro and Hanako who got married yesterday are students of U B C (223) S NP VP NP ga UBC no gakusei da kinoo NP kekkon- kinoo NP kekkon-Hanako to Taroo to 73 S. NP UBC no I gakusei da S NP UBC no \. | gakusei da Taroo Adv NP V Hanako Adv NP V kinoo Hanako kekkon- s i - t a , kinoo Taroo kekkon- s i - t a . Note that i n the d e r i v a t i o n of (223) from (224) R e l a t i v i z a t i o n and Conjunction Reduction are ap p l i e d . Symmetric predicates include not only verbs but als o a d j e c t i v e s and nouns, so Reciprocal Transformation can apply to the underlying stru c t u r e (225) to y i e l d (226). (225) CqTaroo ga Ziroo to kyoodai da^ ] f~ rZiroo ga Taroo to kyoodai (226) Taroo to Ziroo ga kyoodai da. 'Taro and J i r o are brothers. 1 However, the r u l e should not apply to structures l i k e (227)-(230). da.] AND] (227) [ g£ sTaroo ga Hanako AND] (228) [~S C gTaroo ga Ziroo love Taroo o a i - s i - t a Taroo to naka no terms i i kyoodai da] AND good 74 (229) £ s [ gTaroo ga kinoo Hanako to kekkon-si-ta]£gHanako ga se n z i t u kekkon-si-ta] AND ] the other day (230) Q fgTaroo ga Hanako to kekkon-si-ta]£gHanako ga Ziroo to kekkon-si-ta] AND ] The underlined parts i n each stru c t u r e v i o l a t e the conditions stated i n ( 2 2 1 ) . As the verb a i - s i 'to love' i s not a symmetric predicate, (227) v i o l a t e s c o n d i t i o n ( i ) . Structures ( 2 2 8 ) - ( 2 2 9 ) and (230) v i o l a t e conditions ( i i ) and ( i i i ) r e s p e c t i v e l y . 2 . 4 . 2 . 2 . Sentence (5/a) i s derived from (231) through Reciprocal Transformation. (5/a) Basu to torakku to t a k u s i i ga s y o o t o t u - s i - t a . 'The bus, the truck, and the t a x i c o l l i d e d (with one another).' (231) f s |Tgbasu ga torakku to syoototu-si-ta^][*„torakku ga basu to sy 0 0 1 o tu- s i -1 a/] [* torakku ga t a k u s i i to sy 0 0 to t u - s i - t a ^ ] £gtakusii ga torakku to s y o o t o t u - s i - t a j ] f " g t a k u s i i ga basu to sy o o t o t u - s i - t a ^ f g b a s u ga t a k u s i i to s y o o t o t u - s i - t a ^ ] AND] I consider that the a c c e p t a b i l i t y d i f f e r e n c e between sentences (101) and (104) depends on whether t h e i r predicates express a t r a n s i t i v e r e l a t i o n or not. (101) Basu to torakku to t a k u s i i ga tagai n i s y o o t o t u - s i - t a . 'The bus, the truck, and the t a x i c o l l i d e d with one another.* (104) xTaroo to Ziroo to Saburoo ga t a g a i n i kyoodai da. 75 A .transitive r e l a t i o n i s a r e l a t i o n R such that i f xRy and yRz are t r u e , then xRz i s a l s o true. The noun kyoodai 'a brother' expresses a t r a n s i t i v e r e l a t i o n , because i f sentences (232a-b) are t r u e , sentence (232c) i s a l s o true. (232) a. Taroo ga Ziroo to kyoodai da. 'Taro i s a brother of J i r o . ' b. Ziroo ga Saburoo to kyoodai da. ' J i r o i s a brother of Saburo.• c. Taroo ga Saburoo to kyoodai da. •Taro i s a brother of Saburo.* However, the verb syoototu-si 'to c o l l i d e ' does not express such a r e l a t i o n , because the f a c t that sentences (233a-b) are true does not imply that sentence (233c) i s a l s o true. (233) a. Basu ga torakku to s y o o t o t u - s i - t a . •The bus c o l l i d e d with the truck.' b. Torakku ga t a k u s i i to s y o o t o t u - s i - t a . 'The truck c o l l i d e d with the t a x i . ' c. Basu ga t a k u s i i to s y o o t o t u - s i - t a . •The bus c o l l i d e d with the t a x i . * Therefore, we can conclude that t a g a i n i 'each other' can occur i n sentences i n which two or more NP's are conjoined and c o n s t i t u t e a subject, only i f the predicates of these sentences are symmetric predicates which do not express a t r a n s i t i v e r e l a t i o n . 2.4.2.3. I have stated i n sec. 2.3.4 that Conjunct Movement applies on the l a s t c y c l e a f t e r Conjunction Reduction. Though sentences (3/a) and (3/b) are derived by Reciprocal Transformation and Conjunction 76 Reduction r e s p e c t i v e l y , the sentences have the same s t r u c t u r e ; so i f Reciprocal Transformation i s ordered before Conjunct Movement, the l a t t e r r u l e can apply to sentence (3/a) to y i e l d sentence (1/a). However, such a d e r i v a t i o n must be blocked, f o r sentence (1/a) i s considered to be generated i n the base component. Therefore, to prevent Conjunct Movement from applying to sentences l i k e (3/a) which i s derived by Reciprocal Transformation, I assume that Reciprocal Transformation i s ordered a f t e r Conjunct Movement. 77 Chapter-3 CONCLUDING REMARKS 3.1. In Chapter 1, I have made d e t a i l e d examinations of Okutsu (1967), Kuno (1967-68) and Inoue (1976b). Notice, however, that the l a t t e r two works give no e x p l i c i t formulations of the r u l e s i n question. In Chapter 2, I have shown that sentences with conjoined NP's are a l l transformationally derived, and have proposed several transformations f o r the der i v a t i o n s of such sentences. The proposed transformations are Conjunction Reduction, Issyo n i I n s e r t i o n , NP Scrambling i n NP, Conjunct Movement, and Reciprocal Transformation. 3.2. The grammars of Okutsu, Kuno and Inoue presented several problems. Let us examine b r i e f l y how these d i f f i c u l t i e s are overcome i n my grammar. Okutsu's grammar (and perhaps, Inoue's as well) cannot generate sentence (73/a), though Kuno's can. (73/a) Taroo to Hanako ga kekkon-si-tagat-ta. 'Taro and Hanako wanted to get married (to each other).' In my grammar, t h i s sentence i s derived from the underlying s t r u c t u r e (234) i n the f o l l o w i n g manner: 78 (234) AND Taroo ga NP tagat-ta a ' y Hanako ga NP V Taroo to kekkon-si_, 1 Hanako to kekkon-si. I (235) a. [ Q Taroo ga [ y p \_ypHanako to k e k k o n - s i / l l y t a g a t - t a 11 ~) J0 . [ s Hanako ga [ y p L* v pTaroo to kekkon-si.^ f y t a g a t - t a ] 1 ]AND~| D» C s [ s C s Taroo ga Hanako to k e k k o n - s i ^ f g Hanako ga Taroo to kekkon-sijl AND]f ytagat-ta] ] c. f„ Taroo to Hanako ga ke kkon- s i ,"| f v t a g a t - ta") 1 s s0 On the S 1 and S 2 cycles Equi NP Deletion applies to (234) to y i e l d (235a). On the S Q cycle Conjunction Reduction applies to (235a) to y i e l d (235b). Then Reciprocal Transformation applies to (235b) to y i e l d (235c), to which Predicate R a i s i n g applies to derive the sur-face s t r u c t u r e of sentence (73/a). Sentence (29) can be generated by Okutsu*s grammar (and perhaps Inoue's a l s o ) , but i t cannot by Kuno's. (29) Taroo to Hanako ga issyo n i kekkon-si-ta. 'Taro and Hanako got married (to someone) together.* Though Issyo ni Insertion must be accompanied by Conjunct Movement 79 i n Kuno's grammar, the l a t t e r r u l e i s not applied i n the d e r i v a t i o n of sentence (29); so the grammaticality of the sentence cannot be accounted f o r by Kuno's. In my grammar, not only sentence (29) but also sentences (1/b), (3/b) and (15) are derived from the same underlying structure (208) as shown i n sec. 2.4.1. On the other hand, sentence (87) i n the sense of "those who got married to each other"can be generated by nei t h e r Kuno's grammar nor Okutsu's (nor Inoue's perhaps). (87) Kinoo kekkon-si-ta Taroo to Hanako wa UBC no gakusei da. 'Taro and Hanako who got married yesterday are students of U B C In my grammar, t h i s sentence i s derived by Reciprocal Transformation as shown i n sec. 2.4.2.1. Sentence (236) i s semantically anomalous i n the sense that Taro, J i r o and Hanako got married to one another, but Inoue's grammar as w e l l as Kuno's cannot account f o r t h i s semantic anomaly because the conjoined NP's are generated i n the base component i n both grammars. (236) Taroo to Ziroo to Hanako ga kekkon-si-ta. 'Taro, J i r o , and Hanako got married.' In my grammar t h i s sentence i s derived from conjoined sentences, one of which i s (237). (237) Taroo ga Ziroo to kekkon-si-ta. •Taro got married to J i r o . ' This sentence demonstrates why sentence (236) i s semantically anomalous, for sentence (237) implies that the marriage of persons of the same sex was permitted. 81 FOOTNOTES AJapanese examples are transcribed i n the National Romanir za t i o n System ( i . e . , K u n r e i s i k i ) with the f o l l o w i n g modifications: the long vowels ar e • i n d i c a t e d by r e p e t i t i o n of the same vowels, and hyphenation as w e l l as spacing i s used to represent a sequence of morphemes. Each of the examples i s followed by an E n g l i s h trans-l a t i o n and (sometimes) word-by-word glosses. I f an example has several i n t e r p r e t a t i o n s , i t s t r a n s l a t i o n r e f l e c t s the most n a t u r a l one, as long as i t i s not under d i s c u s s i o n . 2 The copula da has two s t y l i s t i c v a r i a n t s : de-ar-u and desu. 3 That i s to say, state predicates are predicates other than verbs. 4 In t h i s sentence, Taroo to Hanako i s a constituent. As any phrase can occupy any p o s i t i o n i n Japanese, we can move Taroo to i n sentence ( i ) to the. i n i t i a l , p o s i t i o n , to y i e l d , a.sentence i d e n t i c a l to ( 3 ) . ( i ) Hanako ga Taroo to kekkon-si-ta. (=(10)) 'Hanako married Taro.• However, Taroo to Hanako i s not a constituent i n the sentence derived from ( i ) . To show t h i s kind. of., information I use a s l a s h (/) i n the f o l l o w i n g manner: ( i i ) Taroo to / Hanako. ga kekkon-si-ta. 'To Taro, Hanako got married.* There i s a s l i g h t pause at the place where the s l a s h i s placed. ^In Japanese, momentary (or punctual) verbs, which i n d i c a t e a momentary a c t i o n , express the continuation of a completed a c t i o n when followed by t e - i - r u . Therefore, as i t 'to go' i s regarded as one of them, i t - t e - i - r u means.something l i k e ( i ) not ( i i ) . ( i ) (Someone) went, (there.) and i s ( s t i l l , t h ere). ( i i ) (Someone) i s going (there). Thus the l i t e r a l t r a n s l a t i o n of sentence (45a) i s something l i k e t h i s : ( i i i ) Tomoko and Yoshiko went to America and are there together now. c Sentences preceded by a question mark i n d i c a t e that native speakers cannot make a uniform judgment as to t h e i r grammaticality. 7 The p a r t i c l e to corresponds to and i n E n g l i s h . Sentences with and are discussed i n a number of works. See Gleitman (1965), Lakoff and Peters (1966), McCawley (1968), Smith (1969), Dougherty (1970, 1971), and Sto.ckwell et a l . (1973). 82 8 The formulations of t h i s r u l e and r u l e (68) are from,. Okutsu (1971). Note that the second Con.j i n both SC's ( S t r u c t u r a l Changes) i s o p t i o n a l l y deleted at a l a t e r stage of the d e r i v a t i o n . q The symbols X, Y and Z are v a r i a b l e s which range over a l l s t r i n g s , i n c l u d i n g the n u l l s t r i n g . x^Kuno assumes that there i s a u n i v e r s a l p r i n c i p l e which converts structures l i k e t h i s i n t o NP NP NP NP. to . . . NP to In a d d i t i o n , he claims that the l a s t to of the conjoined stru c t u r e i s deleted by some transformational r u l e . X AThe symbol + i s used to denote s i s t e r adjunction. Under t h i s adjunction, a phrase marker l i k e ( i ) i s converted i n t o ( i i ) . ( i ) S ( i i ) S That i s , when the element B i s s i s t e r - a d j o i n e d to the l e f t of the node A, B becomes immediately dominated by the node which immediately dominates A. 12 This verb takes a verb s e n t e n t i a l complement. For f u r t h e r d e t a i l , see Nakau (1971). 13 I omit an i r r e l e v a n t d e t a i l i n t h i s s t r u c t u r e . 14 This c o n s t r a i n t i s often r e f e r r e d to as the "Coordinate Structure-Constraint". Ross (1967: 89) defines i t as f o l l o w s : ( i ) The Coordinate Structure Constraint In a coordinate s t r u c t u r e , no conjunct may be moved, nor may any element contained i n a conjunct be moved out of that conjunct. 15 A grammar of Japanese contains a r u l e schema of the form ( i ) S-»S n AND, n£2 where ( i ) represents an i n f i n i t e set of r u l e s , each having the form S —> S S • • • S S AND. These r u l e s generate base structures of the form 83 ( i i ) ' s s S S, AND There i s a u n i v e r s a l p r i n c i p l e which converts s t r u c t u r e s l i k e ( i i ) to ( i i i ) . ( i i i ) AND We have an o b l i g a t o r y r u l e which deletes the l a s t AND. In a d d i t i o n , the other AND's can be deleted by some o p t i o n a l r u l e . Note that AND i s not a r e a l l e x i c a l item. I t i s r e a l i z e d as te i f i t i s dominated by S, and as to and te i f i t happens to be dominated by NP and VP r e s p e c t i v e l y by means of Conjunction Reduction. ^The symbol § i s used to denote Chomsky adjunction. Under t h i s adjunction, a phrase marker l i k e ( i ) i s converted i n t o ( i i ) . ( i ) S ( i i ) S W - A -That . i s , when the element B i s Chomsky-adjoined to the r i g h t of the node A, the o l d node A keeps i t s constituents and a new node A i s created to dominate the old,node A and the element B. 17 I take the p o s i t i o n that a Japanese grammar contains the category VP. This p o s i t i o n i s argued against by Inoue (1976a), who considers that there are no transformations whose SD's r e f e r to VP. However, Inoue's argument i s weak because i t i s based on the obser-va t i o n of the f o l l o w i n g p a i r s of. sentences: ( i ) a. Hanako ga gakkoo de eigo o n a r a t - t a . at English l e a r n 'Hanako learned English at school.' b. Taroo mo s o o - s i - t a . a l s o so do 'Taro d i d so too.' ( i i ) a. Watakusi wa nairon-burausu o te de arai-masu. nylon blouse hand by wash 'I wash a nylon blouse by hand.' b. Watakusi wa ke no seetaa mo soo-si-masu. wool sweater ' ( L i t . ) I do a wool sweater too.' According to her, the phrases gakkoo de i n ( i a ) and nairon-burausu 84 i n ( i i a ) are outside VP and i n s i d e VP r e s p e c t i v e l y . Therefore, the f a c t that the underlined parts i n (a) sentences of ( i ) and ( i i ) are replaced by the pro-expression s o o - s i shows that the replaced phrases are not VP's. Prom t h i s f a c t , Inoue concludes that the SD of the transformation which replaces c e r t a i n phrases by s o o - s i does not r e f e r to VP. However, compare sentences, ( i a ) and ( i i a ) with the f o l l o w i n g sentences: ( i ) c. Taroo wa kyookai de s o o - s i - t a . church •Taro d i d so at church.' ( i i ) c. Watakusi mo soo-si-masu. • I do so too. 1 In these (c) sentences, eigo o narat and nairon-burausu o te de a r a i i n (a) sentences are replaced by s o o - s i r e s p e c t i v e l y . That i s , VP's i n (a) sentences are replaced by the pro-expression i n (c) sentences. This i n d i c a t e s that the SD of t h i s replacement transformation r e f e r s to the category VP. What are needed to describe sentences l i k e ( l b ) and ( l i b ) are simply conditions which make the transformation also a p p l i c a b l e to phrases other than VP. Thus, those (b) sentences do not lead us to the conclusion that the category VP i s not necessary i n a Japanese grammar. 18 I have ignored the problems of tense i n t h i s t h e s i s . 19 Although I take the p o s i t i o n that t h i s r u l e a p p l i e s to (113), one could als o p o s i t that the r u l e applies to (107). In t h i s case, the structure (107) i s changed i n t o ( i ) . ( i ) AND Taroo ga kekkon-s i - t a Hanako ga kekkon-s i - t a 20„ This sentence i s ungrammatical i f the h o n o r i f i c marker o i s i n t e r p r e t e d as modifying imo as w e l l as kome. 21 I assume that a grammar of Japanese contains ( i ) as one of the phrase str u c t u r e r u l e s . ( i ) NP-»NP P However, f o r the sake of s i m p l i c i t y of the tree diagram, I have not used the category symbol P, which stands f o r P a r t i c l e . The dummy symbol which the node P immediately dominates i s not s u b s t i t u t e d by a l e x i c a l transformation; r a t h e r , i t i s s u b s t i t u t e d by Case Marking (such as Subject Marking). Thus, sentence ( i i ) i s derived from an underlying st r u c t u r e l i k e ( i i i ) by Subject Marking. ( i i ) Taroo ga k i - t a . 'Taro came.' come 85 ( i i i ) NP VP NP I Taroo Subject Marking, the formulation of which i s something l i k e ( i v ) , applies to ( i i i ) to s u b s t i t u t e ga f o r the dummy symbol ( i . e . , A | . ( i v ) Subject Marking SD: NP - P - X 1 2 3 SC: 1 ga 3 22. OBLIG (OBLIG=Obligatory) I do not intend to claim that t h i s s t r u c t u r e i s not accept-able; I simply want to say that Conjunction Reduction should not apply to (132). Therefore, sentences l i k e ( i ) are p e r f e c t l y a l l r i g h t . ( i ) Watakusi wa muzukasiku-te tumaranai hon o yon-da. d i f f i c u l t boring read " I read the d i f f i c u l t , boring book.' The underlined part of t h i s sentence has the f o l l o w i n g s t r u c t u r e , but i t i s not derived from (132). 23 This statement implies that a l l NP's can be moved by Pseudo-c l e f t Sentence Formation, as long as the A-over-A p r i n c i p l e i s accepted. This p r i n c i p l e asserts that a l l transformations which r e f e r to A must apply to the topmost instance of A, not dominated A. That i s , i n a phrase marker l i k e ( i ) the transformations apply to A^ not Ag. ( i ) A, Therefore, the ungrammaticality of the f o l l o w i n g sentences, i n which the NP's Taroo to Hanako ga and kuruma o i n sentence (137) are moved by Pseudo-cleft Sentence, Formation, i s due to the idiosyncrasy of each of the p a r t i c l e s attached to the NP's. ( i i ) a. ^Issyo n i kuruma o k a t - t a no wa Taroo to Hanako ga da. b. Taroo to Hanako ga issyo n i kat-ta no wa kuruma o da. That i s , the subject marker _ga and the d i r e c t object marker o_ must be deleted when they are placed before the copula da. On the other hand, the i n d i r e c t object marker n i must not be deleted and the d e l e t i o n of the place marker de_ i s o p t i o n a l . Compare the sentences 86 i n ( i i ) with the f o l l o w i n g p a i r s of sentences: ( i i i ) a. Taroo ga tegami o k a i - t a no wa Hanako n i da. l e t t e r write 'The person to whom Taro wrote the l e t t e r i s Hanako.' b. Taroo ga tegami o k a i - t a no wa Hanako da. ( i v ) a. Taroo ga nooto o k a t - t a no wa ano hon-ya de da. notebook bookstore 'The place where Taro bought a notebook i s that bookstore.' b. Taroo ga nooto o k a t - t a no wa ano hon-ya da. 'The place where Taro bought a notebook i s that bookstore.' The sentences, i n ( i i i ) and ( i v ) are considered to be derived from sentences ( i i i c ) and ( i v c ) r e s p e c t i v e l y through Pseudo-cleft Sentence Formation. ( i i i ) c. Taroo ga tegami o Hanako n i k a i - t a . Taro wrote the l e t t e r to Hanako.' ( i v ) c. Taroo ga ano hon-ya de nooto o k a t - t a . 'Taro bought a notebook at that bookstore.' We have the f o l l o w i n g sentence, i n which kyuu n i seems to be t o p i c a l i z e d . ( i ) Kyuu n i wa d e k i - n a i . 'I cannot do i t i n a hurry.' can not Notice that the phrase kyuu n i i s used with a negative i n sentence ( i ) . As the t r a n s l a t i o n shows, sentence ( i ) i s an abridged sentence. I t s unabridged counterpart would be something l i k e t h i s : ( i i ) Watakusi n i . wa sore ga kyuu n i wa d e k i - n a i . i t 'I cannot do i t i n a hurry.' In t h i s sentence, watakusi n i not kyuu n i i s c l e a r l y the t o p i c . The p a r t i c l e wa which follows kyuu n i i s used to i n d i c a t e the scope of negation. 25 One might claim that (177b) also meets the SD of t h i s r u l e , but t h i s i s not the case. As Hanako to and issyo n i are dominated by Adv i n (177b), the A-over-A p r i n c i p l e stated i n footnote 23 applies to prevent Scrambling from applying to those phrases. 26 This transformation creates a new node. Notice that the node Adv which dominates issyo n i i n (188) i s created by t h i s transformation. Thus one might consider that transformations can perform various operations; however, the operations which transformations can perform are rather r e s t r i c t e d . The f o l l o w i n g are the only possible operations. ( i ) adjunction ( i i ) s u b s t i t u t i o n ( i i i ) d e l e t i o n They are often r e f e r r e d to as elementary transformations. Conjunction Reduction stated i n (106) als o creates a new node. However, i n t h i s case, i t i s not e n t i r e l y new but i s a copy of an o l d node. On the 87 other hand, the node created by Issyo n i I n s e r t i o n i s e n t i r e l y new. Prom t h i s f a c t , i t f o l l o w s that t h i s transformation i s an exceptional one as i t stands. Therefore, to accept such a transformation, i t w i l l be necessary to extend the c l a s s of possible operations of t r a n s f o r -mations. However, i f the node Adv i n question i s not created but i s generated i n the base component, such an extention w i l l not be .neces-sary. In, t h i s case, Issyo n i I n s e r t i o n would have to be formulated i n t h i s way: ( i v ) Issyo n i I n s e r t i o n SD: I - [ N p t H p I AND] P] - Adv - X - V 0PT 1 2 3 4 5 > SC: 1 2 issyo n i 4 5 Condition: 5 i s a marked element. Issyo n i I n s e r t i o n stated above c o n s i s t s of an elementary s u b s t i t u t i o n transformation, while that stated i n (187) c o n s i s t s of an elementary adjunction transformation. I t should be noted that whichever formu-l a t i o n we may adopt, the attached c o n d i t i o n i s necessary to block the d e r i v a t i o n of (29) from sentence (3/c). 27 As an a l t e r n a t i v e to an assumption l i k e mine, we could p o s i t some kind of i n t e r p r e t a t i o n r u l e to account f o r the meaning di f f e r e n c e i n sentence (137). This p o s i t i o n corresponds to the one Akmajian (1973) proposed with r e l a t i o n to sentence (193). Which of those a l t e r n a t i v e s i s c orrect should be subject to e m p i r i c a l t e s t s . 88 BIBLIOGRAPHY Akmajian, A. (1973). "The Role of Focus i n the I n t e r p r e t a t i o n of Anaphoric Expressions." In S. R. Anderson and P. Kiparsky (eds.), A F e s t s c h r i f t f o r Morris H a l l e . New York: H o l t , Rinehart and Winston, pp. 215-226. Bach, E. and R. T. Harms (eds.) (1968). Universal i n L i n g u i s t i c  Theory. New York: H o l t , Rinehart and Winston. B i n n i c k , R. I . , A. Davison, G. M. Green, and J . L. Morgan (eds.) (1969). Papers from the F i f t h Regional Meeting of the Chicago L i n g u i s t i c  Society. Chicago: Department of L i n g u i s t i c s , U n i v e r s i t y of Chicago. Carnap, R. (1958). Introduction to Symbolic Logic and I t s A p p l i c a t i o n s . New York: Dover P u b l i c a t i o n s . Chomsky, N. (1957). Syntactic Structures. The Hague: Mouton & Co. Chomsky, N. (1965). Aspects of the Theory of Syntax. Cambridge, Mass.: The M.I.T. Press. Dougherty, R. (1970). "A Grammar of Coordinate Conjoined Structures: I. " Language 46, pp. 850-898. Dougherty, R. (1971). "A Grammar of Coordinate Conjoined Structures: I I . " Language 47, pp. 298-339. Gleitman, L. R. (1965). "Coordinating Conjunctions i n E n g l i s h . w  Language 41, pp. 260-293. Reprinted i n D. A. Reibel and S. A. Schane (eds.) (1969). Inoue, K. (1970). "Review of Japanese L i n g u i s t i c s : Grammar (Transfor-mational Theory and American L i n g u i s t i c s i n General)." Kokugogaku 81, pp. 55-67. Inoue, K. (1976a). Henkei Bunpoo to Nihongo. V o l . I. Tokyo: Taishuukan. Inoue, K. (1976b). Henkei Bunpoo to Nihongo. V o l . I I . Tokyo: Taishuukan. Jacobs, R. A. and P. S. Rosenbaum (1968). E n g l i s h Transformational  Grammar. Waltham, Mass.: B l a i s d e l l . K a j i t a , M. (1974). "Henkei Bunpoo." In A. Ota (ed.), Outline of  English L i n g u i s t i c s , V o l . IV. Tokyo: Taishuukan. pp. 163-647. Kimball, J . P. (1973). The Formal Theory of Grammar. Englewood C l i f f s , New Jersey: P r e n t i c e - H a l l . 89 Kokuritsu Kokugo Kenkyuujo (1951). Gendaigo no Josh i • Jodooshi. Tokyo: Shuuei Shuppan. Kuno, S. (1967-68). "And to To, N i , Ya." Kotoba no Uchuu, December 1967, January and February 1968. Tokyo: TEC. Kuno, S. (1973). The Structure of Japanese Language. Cambridge, Mass.: M.I.T. Press. 1 Lakoff, G. and S. Peters (1966). "Phrasal Conjunction and Symmetric Predicates." Mathematical L i n g u i s t i c s and Automatic T r a n s l a t i o n , Report No. NSF-17 to The National Science Foundation, Harvard U n i v e r s i t y , Computation Laboratory. Reprinted i n D. A. Reibel and S. A. Schane (eds.) (1969). Langendoen, D. T. (1969). The Study of Syntax. New York: Holt, Rinehart and Winston. Lees, R. B. and E. S. Klima (1963). "Rules f o r E n g l i s h Pronominali-z a t i o n . " Language 39, pp. 17-28. Reprinted i n D. A. R e i b e l and S. A. Schane (eds.) (1969). McCawley, J . D. (1968). "The Role of Semantics i n a Grammar." In E. Bach and R. T. Harms (eds.), Universals i n L i n g u i s t i c Theory, pp. 124-169. Nakau, M. (1971). "Toward a Theory of S e n t e n t i a l Complementation i n Japanese." Eigogaku. Tokyo: Kaitakusha. pp. 104-162. Okutsu, K. (1967). "Taisyoo Kankei Koozoo to Sono Tenkei." Nihongo Kenkyuu. Tokyo: I n t e r n a t i o n a l C h r i s t i a n U n i v e r s i t y , pp. 1-21. Okutsu, K. (1971). Lectures at the I n t e r n a t i o n a l C h r i s t i a n U n i v e r s i t y . Ota, A. (ed.) (1974). Outline of English L i n g u i s t i c s . V o l . IV. Tokyo: Taishuukan. R e i b e l , D. A. and S. A. Schane (eds.) (1969). Modern Studies i n Eng l i s h : Readings i n Transformational Grammar. Englewood C l i f f s , New Jersey: P r e n t i c e - H a l l . Ross, J . R. (1967). "Constraints on Variables i n Syntax." Unpublished Doctoral d i s s e r t a t i o n , M.I.T. Reproduced by the Indiana U n i v e r s i t y L i n g u i s t i c s Club (1968). (Page references are to t h i s version.) Ross, J . R. (1969). "Guess Who?" In R. I. Binnick et a l . (eds.), Papers from the F i f t h Regional Meeting of the Chicago L i n g u i s t i c  Society, pp. 252-286. Smith, C. S. (1969). "Ambiguous Sentences with AND." In D. A. R e i b e l and S. A. Schane (eds.). 90 Stockwell, R. P.,-P. Schachter, and B. H. Partee (1973). The Major  Syntactic Structures of English. New York: H o l t , Rinehart and Winston. Yasui, M. (ed.) (1971). Kenkyusha's Dictionary of New L i n g u i s t i c s . Tokyo: Kenkyusha. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0094169/manifest

Comment

Related Items